對于交易成本，更進幾步說，有必要去弄清楚利潤是否僅僅是對因在特定時期內持有風險資所承擔風險的補償。我們一共申請787用于計算對象與51個市場指數并且利用計要前期機技術進行跟蹤，關于交易成本(參見章節2.3和3.3的第三章和附錄B技術交易規則parameterizations)。我們檢測的有三種不同交易案數據。如果我們是本地的商人,可以運用我們的技術交易來規范當地貨幣,我們所表達的利潤是用本地幣表示的。如果沒有在股票市場交易，是可以獲得當地無風險利率持有股。然而由于折舊率可能影響利潤所以當地貨幣美元需要重新計算。

 

我們請參考3.4節討論如何計算道瓊斯利潤的技術交易規則。如果0和0.25%假設為每股實現貿易成本率,交易案3表5.3和5.4顯示為每個主要股票市場指數的統計數據，我們可以做為參考從而選擇的最佳策略。列2顯示了參數的最佳策略。對于平均機會成本策略表格中所反應的這些參數”(短期、長期)”加上細化參數”可以設立很好的止損點。在交易區間的情況下,也稱為支持與阻力(SR)策略,相關參數”(當地最大和最小的天數計算)”加上優化參數與移動平均線可以做為企業經營的目標。

 

Transaction costs. Furthermore, it is necessary to test whether the pro?ts are just the compensation for bearing the risk of holding the risky asset during certain periods. In total we apply 787 objective computerized trend-following technical trading techniques with and without transaction costs to the 51 market indices (see sections 2.3 and 3.3 and Appendix B of Chapter 3 for the technical trading rule parameterizations). We consider three di?erent trading cases. First we do as if we are a local trader and we apply our technical trading rule set to the indices expressed in local currency and we compute the pro?ts expressed in local currency. If no trading position in the stock market index is held, then the local risk-free interest rate is earned. Due to depreciation however, it is possible that pro?ts in local currencies disappear when recomputed in US Dollars. Therefore we also consider the problem from the perspective of an US-based trader. Trading signals are then generated in two di?erent ways: ?rstly on the indices expressed in local currency and secondly on the indices recomputed in US Dollars. Recomputation of local indices in US Dollars is done to correct for possible trends in the levels of stock market indices caused by a declining or advancing exchange rate of the local currency against the US Dollar. If the US-based trader holds no trading position in the stock market index, then the US risk-free interest rate is earned. Summarized we examine the following trading cases: Trader Trading case 2 US trader Trading case 3 US trader Index in Pro?ts in。

 

We refer to section 3.4 for a discussion on how the technical trading rule pro?ts are computed. If 0 and 0.25% costs per trade are implemented, then for trading case 3 tables 5.3 and 5.4 show for each local main stock market index some statistics of the best strategy selected by the mean return criterion. Column 2 shows the parameters of the best strategy. In the case of a moving-average (MA) strategy these parameters are “[short run MA, long run MA]” plus the re?nement parameters “[%-band ?lter, time delay ?lter, ?xed holding period, stop-loss]”. In the case of a trading range break, also called support-and-resistance (SR), strategy, the parameters are “[the number of days over which the local maximum and minimum is computed]” plus the re?nement parameters as with the moving averages. In the case of a ?lter (FR) strategy the parameters are “[the %-?lter, time delay ?lter, ?xed holding period]”. 

 

Best-selected strategy over the buy-and-hold benchmark. Column 7 shows the maximum cumulative loss the best strategy generates. Columns 8, 9 and 10 show the number of trades, the percentage of pro?table trades and the percentage of days pro?table trades last. Finally, the last column shows the standard deviation of the returns of the indices during pro?table trades divided by the standard deviation of the returns of the indices during non-pro?table trades. To summarize, for trading case 3 table 5.6A (i.e. table 5.6 panel A) shows for each index the mean yearly excess return over the buy-and-hold benchmark of the best strategy selected by the mean return criterion, after implementing 0, 0.10, 0.25, 0.50, 0.75 and 1% costs per trade. This wide range of costs captures a range of di?erent trader types. For example, ?oor traders and large investors, such as mutual funds, can trade against relatively low transaction costs in the range of 0.10 to 0.25%. Home investors face higher costs in the range of 0.25 to 0.75%, depending whether they trade through the internet, by telephone or through their personal account manager. Next, because of the bid-ask spread, extra costs over the transaction costs are faced. By examining a wide range of 0 to 1% costs per trade, we belief that we can capture most of the cost possibilities faced in reality by most of the traders. At the bottom of table 5.6A, the row labeled “Average 3” shows for each transaction cost case the average over the results for trading case 3 as presented in the table. For comparison with the other two trading cases the row labeled “Average 1” shows the average over the results if trading case 1 is examined and the row labeled “Average 2” shows the average over the results if trading case 2 is examined. Table 5.6A clearly shows that for each stock market index the best technical trading strategy selected by the mean return criterion is capable of beating the buy-and-hold benchmark, even after correction for transaction costs. If transaction costs increase from 0 to 1% per trade, then it can be seen that the excess returns decline on average from 49.14 to 17.22%. However, even in the large 1% costs per trade case, the best technical trading strategy is superior to the buy-and-hold strategy. The lowest excess returns are found for the West European stock market indices, while the highest excess returns are found for the Asian and Latin American stock market indices. No large di?erences are found between the three trading cases. The results, as summarized by the averages in the bottom rows of table 5.6A, are similar. From table 5.3 it can be seen that in the case of zero transaction costs the best-selected strategies are mainly strategies that generate a lot of signals. Trading positions are held for only a few days. For example, the best technical trading strategy found for the MSCI World Index is a single crossover moving-average rule, with no extra re?nements, which generates a signal when the price series crosses a 2-day moving average. The mean yearly。

 

Return is equal to 52%, which corresponds with a mean yearly excess return of 40%. The Sharpe ratio is equal to 0.1461 and the excess Sharpe ratio is equal to 0.1349. These excess performance measures are considerably large. The maximum loss of the strategy is 18.7%, half less than the maximum loss of buying and holding the MSCI World Index, which is equal to 38.7%. The number of trades is very large, once every 2.4 days, but also the percentage of pro?table trades is very large, namely 74.8%. These pro?table trades span 86% of the total number of trading days. Similar good results are also found for the other stock market indices. For 42 of the 51 indices the maximum loss of the best strategy is less than the largest cumulative loss of the buy-and-hold strategy. For most indices the percentage of pro?table trades is larger than 70% and these pro?table trades span more than 80% of the total number of trading days. Although the Sharpe ratio of the buy-and-hold was negative for 23 indices, indicating that these indices were not able in beating a continuous risk free investment, it is found for all indices that the best-selected strategy shows a positive Sharpe ratio. If transaction costs are increased to 0.25% per trade, then table 5.4 shows that the bestselected strategies are strategies which generate substantially fewer signals in comparison with the zero transaction costs case. Trading positions are now held for a longer period. For example, the best strategy found for the MSCI World Index is a single crossover moving-average rule which generates signals if the price series crosses a 200-day moving average and where the single re?nement is a 2.5%-band ?lter. This strategy generates a trade every 13 months. However due to transaction costs the performance of the technical trading rules decreases and also the percentage of pro?table trades and the percentage of days pro?table trades last decreases for most indices in comparison with the zero transaction costs case. However for all indices, the Sharpe ratio of the best strategy is still positive. This continues to be the case even if costs are increased to 1% per trade. Similar results are found for the two other trading cases. CAPM If no transaction costs are implemented, then for trading case 3 it can be seen from the last column in table 5.3 that the standard deviations of the daily returns during pro?table trades are higher than the standard deviations of the daily returns during non-pro?table trades for almost all stock market indices, except for the Indonesian Jakarta Composite, the Finnish HEX, the Swiss SMI and the Irish ISEQ. However, if 0.25% costs per trade are calculated, then for only 24 indices out of 51 the standard deviation ratio is larger than one. Similar results are found for the other two trading cases. According to the e?cient markets hypothesis it is not possible to exploit a data set with past information.

 

To predict future price changes. The good performance of the technical trading rules could therefore be the reward for holding a risky asset needed to attract investors to bear the risk. Since the technical trading rule forecasts only depend on past price history, it seems unlikely that they should result in unusual risk-adjusted pro?ts. To test this hypothesis we regress Sharpe-Lintner capital asset pricing models (CAPMs).

 

Combine these with two di?erent choices for the market portfolio M , we estimated in total six di?erent CAPMs for each index. The coe?cient β measures the riskiness of the active technical trading strategy relatively to the passive strategy of buying and holding the market portfolio. If β is not signi?cantly di?erent from one, then it is said that the strategy has equal risk as a buying and holding the market portfolio. If β > 1 (β < 1), then it is said that the strategy is more risky (less risky) than buying and holding the market portfolio and that it therefore should yield larger (smaller) returns. The coe?cient α measures the excess return of the best strategy applied to stock i after correction of bearing risk. If it is not possible to beat a broad market portfolio after correction for risk and hence technical trading rule pro?ts are just the reward for bearing risk, then α should not be signi?cantly di?erent from zero. We estimated the Sharpe-Lintner CAPMs in the case of 0, 0.10, 0.25, 0.50, 0.75 and 1% costs per trade. For trading case 3 table 5.7 shows in the cases of 0 and 0.50% transaction costs the estimation results if for each index the best strategy is selected by the mean return criterion and if the market portfolio is chosen to be the local main stock market index. Estimation is done with Newey-West (1987) heteroskedasticity and autocorrelation consistent (HAC) standard errors. Table 5.9 summarizes the CAPM estimation results for all trading cases and for all transaction cost cases by showing the number of indices for which signi?cant estimates of α or β are found at the 10% signi?cance level. For example, for the best strategy applied to the MSCI World Index in the case of zero transaction costs, the estimate of α is signi?cantly positive at the 1% signi?cance level and equal to 13.42 basis points per day, that is approximately 33.8% on a yearly basis. The estimate of β is signi?cantly smaller than one at the 10% signi?cance level, which indicates that although the strategy generates a higher reward than simply buying.

 

Table 5.9: Summary: signi?cance CAPM estimates, mean return criterion. For each transaction cost case, the table shows the number of indices for which signi?cant estimates are found at the 10% signi?cance level for the coe?cients in the Sharpe-Lintner CAPM. The local main stock market index is taken to be the market portfolio in the CAPM estimations. Columns 1 and 2 show the number of indices for which the estimate of α is signi?cantly negative and positive. Columns 3 and 4 show the number of indices for which the estimate of β is signi?cantly smaller and larger than one. Column 5 shows the number of indices for which the estimate of α is signi?cantly positive as well as the estimate of β is signi?cantly smaller than one. Column 6 shows the number of indices for which the estimate of α is signi?cantly positive as well as the estimate of β is signi?cantly larger than one. Note that the number of indices analyzed is equal to 51.

 

And holding the index, it is less risky. If transaction costs increase to 1% per trade, then α decreases to 1.82 basis points (4.6% yearly), but still is signi?cantly positive at the 10% signi?cance level. However, the estimate of β is not signi?cantly smaller than one anymore, if as little as 0.25% costs per trade are charged. It becomes even signi?cantly larger than one if 1% transaction costs are implemented, which indicates that the strategy applied to the MSCI World Index is riskier than buying and holding the market index. If the local main stock market index is taken to be the market portfolio in the CAPM estimations and if zero transaction costs are implemented, then, as further can be seen in the tables, also for the other indices the estimate of α is signi?cantly positive at the 10% signi?cance level. Further the estimate of β is signi?cantly smaller than one for 29 indices. For none of the indices the estimate of β is signi?cantly larger than one. The estimate of α in general decreases as costs per trade increases and becomes less signi?cant for more indices. However in the 0.10, 0.25, 0.50, 0.75 and 1% costs per trade cases for.

 

Example, still for respectively 45, 44, 37, 32 and 27 indices out of 51 the estimate of α is signi?cantly positive at the 10% signi?cance level. The estimate of β is signi?cantly smaller than one for 28, 28, 28, 33 and 34 indices, in the 0.10, 0.25, 0.50, 0.75 and 1% costs per trade cases, indicating that even in the presence of high costs, the best selected technical trading strategies are less risky than the buy-and-hold strategy. The number of data series for which the estimate of β is signi?cantly smaller than one increases as transaction costs increase. This is mainly caused because as transaction costs increase, by the selection criteria strategies are selected which trade less frequently and are thus less risky. Notice that for a large number of cases it is found that the estimate of α is signi?cantly positive while simultaneously the estimate of β is signi?cantly smaller than one. This means that the best-selected strategy did not only generate a statistically signi?cant excess return over the buy-and-hold benchmark, but is also signi?cantly less risky than the buy-and-hold benchmark. The results for the two other trading cases are similar. If the MSCI World Index is used as market portfolio in the CAPM estimations, then the results for α become less strong4 . In the case of zero transaction costs for 46 stock market indices it is found that the estimate of α is signi?cantly di?erent from zero. In the 0.10, 0.25, 0.50, 0.75 and 1% costs per trade cases, for respectively 40, 34, 24, 24 and 19 indices out of 51 the estimate of α is signi?cantly positive at the 10% signi?cance level. However still the estimate of β is signi?cantly smaller than one for 41, 41, 40, 40 and 42 indices in the 0.10, 0.25, 0.50, 0.75 and 1% costs per trade case. From these ?ndings we conclude that there are trend-following technical trading techniques which can pro?tably be exploited, even after correction for transaction costs, when applied to local main stock market indices. As transaction costs increase, the best strategies selected are those which trade less frequently. Furthermore, if a correction is made for risk by estimating Sharpe-Lintner CAPMs, then it is found for many local main stock market indices that the best strategy has forecasting power, i.e. α > 0. It is also found that in general the best strategy is less risky, i.e. β < 1, than buying and holding the market portfolio. Hence, for most stock market indices, we can reject the null hypothesis that the pro?ts of technical trading are just the reward for bearing risk. Data snooping The question remains open whether the ?ndings in favour of technical trading for particular indices are the result of chance or of real superior forecasting power. 

 

Apply White’s (2000) Reality Check (RC) and Hansen’s (2001) Superior Predictive Ability (SPA) test. Because Hansen (2001) showed that White’s RC is biased in the direction of one, p-values are computed for both tests to investigate whether these tests lead in some cases to di?erent inferences. In the case of 0 and 0.25% transaction costs table 5.8 shows for trading case 3 the nominal, RC and SPA-test p-values, if the best strategy is selected by the mean return criterion5 . Table 5.10 summarizes the results for all transaction cost cases by showing the number of indices for which the corresponding p-value is smaller than 0.10. That is, the number of data series for which the null hypothesis is rejected at the 10% signi?cance level.

costs 0% 0.10% 0.25% 0.50% 0.75% 1% Trading case 3 pn pW pH 51 8 27 51 6 15 51 2 6 51 0 2 51 0 1 51 0 1.

 

Table 5.10: Summary: Testing for predictive ability, mean return criterion. For each transaction cost case, the table shows the number of indices for which the nominal (pn ), White’s (2000) Reality Check (pW ) or Hansen’s (2001) Superior Predictive Ability test (pH ) p-value is smaller than 0.10. Note that the number of indices analyzed is equal to 51.

 

The nominal p-value, also called data mined p-value, tests the null hypothesis that the best strategy is not superior to the buy-and-hold benchmark, but does not correct for data snooping. From the tables it can be seen that this null hypothesis is rejected for all indices in all cost cases at the 10% signi?cance level. However, if we correct for data snooping, then in the case of zero transaction costs we ?nd for only 8 of the stock market indices that the null hypothesis that the best strategy is not superior to the benchmark after correcting for data snooping is rejected by the RC, while for 27 indices the null hypothesis that none of the alternative strategies is superior to the buy-and-hold benchmark after correcting for data snooping is rejected by the SPA-test. The two data snooping tests thus give contradictory results for 19 indices. Thus the RC misguides the researcher in several cases by not rejecting the null. The number of contradictory results decreases to 9 if 0.10% costs per trade are implemented and to 4, 2, 1 and 1 if 0.25, 0.50, 0.75 and 1% costs per trade are implemented. In the 0.10% costs per trade case, the SPA-test rejects for 15 indices its null hypothesis, but this number declines to 2 in the.

 

Technical trading rule performance Similar to tables 5.3 and 5.4, table 5.5 shows for trading case 3 some statistics of the best strategy selected by the Sharpe ratio criterion, if 0 or 0.25% costs per trade are implemented. Only the results for those indices are shown for which the best strategy selected by the Sharpe ratio criterion di?ers from the best strategy selected by the mean return criterion. Further, to summarize, table 5.6B shows for each index the Sharpe ratio of the best strategy selected by the Sharpe ratio criterion, after implementing 0, 0.10, 0.25, 0.50, 0.75 and 1% costs per trade, in excess of the Sharpe ratio of the buy-and-hold benchmark. For each index and for each transaction costs case it is found that the excess Sharpe ratio is considerably positive. In the last row of table 5.6B it can be seen that on average the excess Sharpe ratio declines from 0.0672 to 0.0320 if transaction costs increase from 0 to 1% per trade. Table 5.5 shows that the best strategies selected in the case of zero transaction costs are mainly strategies which trade frequently. For most indices, except 10, the best-selected strategy is the same as in the case that the best strategy is selected by the mean return criterion. If costs are increased to 0.25%, then the best strategies selected are those which trade less frequently. Now for 22 indices the best-selected strategy di?ers from the case when the best strategy is selected by the mean return criterion. The results for the two other trading cases are similar. As for the mean return criterion it is found that for each stock market index the best technical trading strategy, selected by the Sharpe ratio criterion, beats the buy-and-hold benchmark and that this strategy can pro?tably be exploited, even after correction for transaction costs. CAPM The estimation results of the Sharpe-Lintner CAPM shown in tables 5.7B and 5.11 for the Sharpe ratio selection criterion are similar to the estimation results shown in tables 5.7A and 5.9 for the mean return selection criterion. If zero transaction costs are implemented, then it is found for trading case 3 that for all 51 indices the estimate of α is signi?cantly.

 

Positive at the 10% signi?cance level. This number decreases from 37 to 28 if transaction costs increase from 0.50 to 1% per trade. As for the mean return selection criterion, for many indices it is found that the estimate of α is signi?cantly positive and that simultaneously the estimate of β is signi?cantly smaller than one. Thus, after correction for transaction costs and risk, for more than half of the indices it is found that the best technical trading strategy selected by the Sharpe ratio criterion signi?cantly outperforms the buy-and-hold benchmark strategy and is even signi?cantly less risky. If the MSCI World Index is taken to be the market portfolio, then the results for α become less strong, as in the mean return selection criterion case. If transaction costs increase from 0% to 0.50 and 1%, then the number of indices for which a signi?cant estimate of α is found declines from 46 to 25 and 18.

 

Table 5.11: Summary: signi?cance CAPM estimates, Sharpe ratio criterion. For each transaction cost case, the table shows the number of indices for which signi?cant estimates are found at the 10% signi?cance level for the coe?cients in the Sharpe-Lintner CAPM. The local main stock market index is taken to be the market portfolio in the CAPM estimations. Columns 1 and 2 show the number of indices for which the estimate of α is signi?cantly negative and positive. Columns 3 and 4 show the number of indices for which the estimate of β is signi?cantly smaller and larger than one. Column 5 shows the number of indices for which the estimate of α is signi?cantly positive as well as the estimate of β is signi?cantly smaller than one. Column 6 shows the number of indices for which the estimate of α is signi?cantly positive as well as the estimate of β is signi?cantly larger than one. Note that the number of indices analyzed is equal to 51.

 

In the case of 0 and 0.25% transaction costs table 5.8B shows for trading case 3 the nominal, White’s RC and Hansen’s SPA-test p-values, if the best strategy is selected by the Sharpe ratio criterion. For trading case 3 table 5.12 summarizes the results for all transaction cost cases by showing the number of indices for which the corresponding p-value is smaller than 0.10.

costs 0% 0.10% 0.25% 0.50% 0.75% 1% Trading case 3 pn pW pH 51 24 35 51 17 28 51 7 23 51 3 16 51 3 15 51 1 13

 

Table 5.12: Summary: Testing for predictive ability, Sharpe ratio criterion. For each transaction cost case, the table shows the number of indices for which the nominal (pn ), White’s (2000) Reality Check (pW ) or Hansen’s (2001) Superior Predictive Ability test (pH ) p-value is smaller than 0.10. Note that the number of indices analyzed is equal to 51.

 

The results for the Sharpe ratio selection criterion di?er from the results for the mean return selection criterion. If the nominal p-value is used to test the null hypothesis that the best strategy is not superior to the benchmark of buy-and-hold, then the null is rejected for all indices at the 10% signi?cance level for all cost cases. If a correction is made for data snooping, then it is found in the case of zero transaction costs that for 24 indices the null hypothesis that the best strategy is not superior to the buy-and-hold benchmark is rejected by the RC. However, for 35 indices the null hypothesis that none of the alternative strategies is superior to the buy-and-hold benchmark after correcting for data snooping is rejected by the SPA-test. Thus for half of the indices we ?nd that the best technical trading rule has forecasting power even when correcting for the speci?cation search. These numbers are higher than for the mean return selection criterion. In total we ?nd for 11 indices contradictory results, which is less than for the mean return selection criterion. Even in the case of 0.10% costs per trade, the number of indices for which the RC and the SPA-test reject the null is high, namely for 17 and 28 indices respectively. However, if transaction costs are increased any further, then the number of indices for which the RC rejects its null declines sharply: to 7, 3, 3, 1 in the 0.25, 0.50, 0.75 and 1% transaction costs cases. In contrast, the SPA-test rejects its null for 23, 16, 15 and 13 indices in the 0.25, 0.50, 0.75 and 1% transaction costs cases. Note that these results di?er substantially from the mean return selection criterion in which case under 0.25, 0.50, 0.75 and 1% costs per trade the null of no superior predictive ability was rejected for.

 

In section 3.7 we argued to apply a recursive out-of-sample forecasting approach to test whether technical trading rules have true out-of-sample forecasting power. For example, recursively at the beginning of each month it is investigated which technical trading rule performed the best in the preceding six months (training period) and this strategy is used to generate trading signals during the coming month (testing period). In this section we apply the recursive out-of-sample forecasting procedure to the main local stock market indices examined in this chapter. We de?ne the training period on day t to last from t − T r until and including t − 1, where T r is the length of the training period. The testing period lasts from t until and including t + T e − 1, where T e is the length of the testing period. At the end of the training period the best strategy is selected by the mean return or Sharpe ratio criterion. Next, the selected technical trading strategy is applied in the testing period to generate trading signals. After the end of the testing period this procedure is repeated again until the end of the data series is reached. For the training and testing periods we use 28 di?erent parameterizations of [T r , T e] which can be found in Appendix B of Chapter 4. If trading case 3 is applied, then in the case of 0.25% transaction costs, tables 5.13 and 5.14 show for the local main stock market indices some statistics of the best recursive optimizing and testing procedure, if the best strategy in the training period is selected by the mean return and Sharpe ratio criterion respectively. Because the longest training period is one year, the results are computed for the period 1983:1-2002:6. Tables 5.15A and 5.15B summarize the results for both selection criteria in the case of 0, 0.10, 0.25, 0.50, 0.75 and 1% costs per trade. In the second to last row of table 5.15A it can be seen that, if in the training period the best strategy is selected by the mean return criterion, then the excess return over the buy-and-hold of the best recursive optimizing and testing procedure is, on average, 37.72, 30.60, 21.41, 12.4, 7.05 and 4.47% yearly in the case of 0, 0.10, 0.25, 0.50, 0.75 and 1% costs per trade. If transaction costs increase, then the best recursive optimizing and testing procedure becomes less pro?table. 

 

After correction for transaction costs, are mainly found for the Asian, Latin American, Middle East and Russian stock market indices. For example, the best recursive optimizing and testing procedure generates an excess return over the buy-and-hold of 43.52, 32.42, 20.99, 12.61, 7.50 and 4.88% yearly for the Argentinean Merval, after implementing 0, 0.10, 0.25, 0.50, 0.75 and 1% transaction costs. However, for the US, Japanese and most Western European stock market indices the recursive out-of-sample forecasting procedure does not show to be pro?table, after implementing transaction costs. If the Sharpe ratio criterion is used for selecting the best strategy during the training period, then the Sharpe ratio of the best recursive optimizing and testing procedure in excess of the Sharpe ratio of the buy-and-hold benchmark is on average 0.0544, 0.0419, 0.0298, 0.0164, 0.0086 and 0.0052 in the case of 0, 0.10, 0.25, 0.50, 0.75 and 1% costs per trade (see second to last row of table 5.15B). As for the mean return selection criterion, the best recursive optimizing and testing procedure does not generate excess Sharpe ratios over the buy-and-hold for the US and most Western European indices in the presence of transaction costs. The best results are mainly found for the Latin American, Egyptian and Asian stock market indices. For comparison, the last rows in tables 5.15A and 5.15B show the average over the results of the best strategies selected by the mean return or Sharpe ratio criterion in sample for each index tabulated. As can be seen, clearly the results of the best strategies selected in sample are better than the results of the best recursive out-of-sample forecasting procedure. For the cases that the best strategy in the optimizing period is selected by the mean return and Sharpe ratio criterion respectively, tables 5.16A and 5.16B show for the 0 and 0.50% transaction cost cases the estimation results of the Sharpe-Lintner CAPM (see equation 5.1), where the return in US Dollars of the best recursive optimizing and testing procedure in excess of the US risk-free interest rate is regressed against a constant α and the return of the local main stock market index in US Dollars in excess of the US risk-free interest rate. Tables 5.17A, B summarize the CAPM estimation results for the two possible choices of the market portfolio and for all transaction cost cases by showing the number of indices for which signi?cant estimates of α or β are found at the 10% signi?cance level. Estimation is done with Newey-West (1987) heteroskedasticity and autocorrelation consistent (HAC) standard errors. If the local main stock market index is taken to be the market portfolio and if the best strategy in the training period is selected by the mean return criterion, then in the case of zero transaction costs it can be seen in table 5.17A that for 37 indices a signi?cantly positive estimate of α is found. As can be seen in table 5.16A, mainly for the US, Japan.

 

Table 5.17: Summary: signi?cance CAPM estimates for best out-of-sample testing procedure. For each transaction cost case, the table shows the number of indices for which signi?cant estimates are found at the 10% signi?cance level for the coe?cients in the Sharpe-Lintner CAPM. In panel A the local main stock market index and in panel B the MSCI World Index is taken to be the market portfolio in the CAPM estimations. Columns 1 and 2 show the number of indices for which the estimate of α is signi?cantly negative and positive. Columns 3 and 4 show the number of indices for which the estimate of β is signi?cantly smaller and larger than one. Column 5 shows the number of indices for which the estimate of α is signi?cantly positive as well as the estimate of β is signi?cantly smaller than one. Column 6 shows the number of indices for which the estimate of α is signi?cantly positive as well as the estimate of β is signi?cantly larger than one. Note that the number of indices analyzed is equal to 51.

 

And Western European countries the estimate of α is neither signi?cantly negative nor positive at the 10% signi?cance level. As transaction costs increase to 0.50%, the number of signi?cant estimates of α decreases to 16. Signi?cant estimates for α are then mainly found for the Asian stock market indices. As transaction costs increase even further to 1%, then the number of signi?cant estimates of α decreases to 7. Signi?cant estimates for α are then found only for the Peru Lima General, Indonesia Jakarta Composite, Pakistan Karachi 100, Sri Lanka CSE All Share, Thailand SET, and the Egypt CMA. If the Sharpe ratio selection criterion is used to select the best strategy in the training period of the recursive optimizing and testing procedure, then the results are similar as for the mean return selection criterion. If transaction costs increase to 1%, then signi?cant estimates of α are found only for the Chile IPSA, Peru Lima General, Sri Lanka CSE All Share, Norway OSE All Share, Russia Moscow Times, and the Egypt CMA. However, if the MSCI World Index is taken to be market portfolio in the CAPM regression, then the results become worse, as can be seen in table 5.17B. In the case of the mean return selection criterion the number of signi?cant estimates of α decreases from 31 to 1 if transaction costs increase from 0 to 1%. Only for the Egypt CMA the estimate of α is signi?cantly positive at the 10% signi?cance level if transaction costs are equal to 1% per trade. If the Sharpe ratio selection criterion is used to select the best strategy in the training period, then also for the Russia Moscow Times the estimate of α is signi?cantly positive at the 10% signi?cance level. Hence, after correction for su?ciently high transaction costs and risk, it can be concluded, independently of the selection criterion used, that the best recursive optimizing and testing procedure shows no statistically signi?cant out-of-sample forecasting power for local main stock market indices world wide. Only for low transaction costs (≤ 0.25% per trade) technical trading shows statistically signi?cant out-of-sample forecasting power for the Asian, Chilean, Czech, Greece, Mexican, Russian and Turkish stock market indices. In contrast, for the US, Japanese and most Western European stock market indices no signi?cant out-of-sample forecasting power is found, even for low transaction costs.

 

Selection criterion over the mean return selection criterion is that it selects the strategy with the highest return/risk pay-o?. Although for 23 stock market indices it is found that they could not even beat a continuous risk free investment, we ?nd for both selection criteria that for each index a technical trading strategy can be selected that is capable of beating the buy-and-hold benchmark, even after correction for transaction costs. The pro?ts generated by the technical trading strategies could be the reward necessary to attract investors to bear the risk of holding the asset. To test this hypothesis we estimate Sharpe-Lintner CAPMs. For each local stock market index the daily return of the best strategy in excess of the risk-free interest rate is regressed against a constant (α) and the daily return of buying and holding a market portfolio in excess of the riskfree interest rate. The coe?cient of the last regression term is called β and measures the riskiness of the strategy relatively to buying and holding the market portfolio. The market portfolio is taken to be the local stock market index, but we also examine the possibility that the market portfolio is represented by the MSCI World Index. If technical trading rules do not generate excess pro?ts after correction for risk, then α should not be signi?cantly di?erent from zero. In the case of zero transaction costs case, it is found for the mean return as well as the Sharpe ratio criterion that for all indices the estimate of α is signi?cantly positive at the 10% signi?cance level, if the local index is used as the market portfolio. Even if transaction costs are increased to 1% per trade, then we ?nd for more than half of the indices that the estimate of α is still signi?cantly positive. Moreover it is found that the estimate of β is simultaneously signi?cantly smaller than one for most indices. Thus for both selection criteria we ?nd for approximately half of the indices that in the presence of transaction costs the best technical trading strategies have forecasting power and even reduce risk. If the MSCI World Index is used as market portfolio in the CAPM estimations, then the results for α become less strong, but even in the 0.50% costs per trade case, for almost half of the indices the estimate of α is signi?cantly positive. An important question is whether the positive results found in favour of technical trading are due to chance or the fact that the best strategy has genuine superior forecasting power over the buy-and-hold benchmark. This is called the danger of data snooping. We apply White’s (2000) Reality Check (RC) and Hansen’s (2001) Superior Predictive Ability (SPA) test, to test the null hypothesis that the best strategy found in a speci?cation search is not superior to the benchmark of a buy-and-hold if a correction is made for data snooping. Hansen (2001) showed that White’s RC is biased in the direction of one, caused by the inclusion of poor strategies.

 

Return selection criterion that the RC and the SPA-test for 19 out of 51 indices lead to di?erent conclusions. The SPA-test ?nds for more than half of the indices that the best strategy does beat the buy-and-hold signi?cantly after correction for data snooping and the inclusion of bad strategies. Thus the biased RC misguides the researcher in several cases by not rejecting the null. However, if as little as 0.25% costs per trade are implemented, then both tests lead for almost all indices to the same conclusion: the best technical trading strategy selected by the mean return criterion is not capable of beating the buy-and-hold benchmark after correcting for the speci?cation search that is used to ?nd the best strategy. In contrast, for the Sharpe ratio selection criterion we ?nd totally di?erent results. The SPA-test rejects the null hypothesis for 35 indices in the case of zero transaction costs, while the RC rejects the null hypothesis for 24 indices. If costs are increased further to even 1% per trade, then for approximately a quarter of the indices analyzed, the SPA-tests rejects the null of no superior predictive ability at the 10% signi?cance level, while the RC rejects the null for only one index. We ?nd for the Sharpe ratio selection criterion large di?erences between the two testing procedures. Thus the inclusion of poor performing strategies, for which is corrected in the SPA-test, can indeed in?uence the inferences about the predictive ability of technical trading rules. Next we apply a recursive optimizing and testing method to test whether the best strategy found in a speci?cation search during a training period also shows forecasting power during a testing period thereafter. For example, every month the best strategy from the last 6 months is selected to generate trading signals during that month. In total we examine 28 di?erent training and testing period combinations. In the case of zero transaction costs, the best recursive optimizing and testing procedure yields on average an excess mean return over the buy-and-hold of 37.72% yearly, if the best strategy in the training period is selected by the mean return criterion. Thus the best strategy found in the past continues to generate good results in the future. If transaction costs increase, then the excess mean returns on average decline. In the presence of 1% transaction costs the excess mean return over the buy-and-hold benchmark is on average 4.47% yearly. For both selection criteria, mainly pro?table results are found for the Asian, Latin American, Middle East and Russian stock market indices. No pro?table results are found for the US, Japanese and Western European stock market indices. However, estimation of SharpeLintner CAPMs indicates that the economic pro?ts of technical trading in almost all stock market indices, except the Egypt CMA and the Russia Moscow Times, can be explained by risk, after a correction is made for su?ciently high transaction costs. Only for transaction costs below or equal to 0.25% some risk-corrected out-of-sample forecasting power is found for the Asian, Latin American, Middle East and Russian stock market indices.

 

Hence, in short, after correcting for su?ciently high transaction costs, risk, datasnooping and out-of-sample forecasting, we conclude that objective trend-following technical trading techniques, applied to local main stock market indices all over the world, are not genuine superior, as suggested by their in-sample performance results, to the buy-andhold benchmark. Only for su?ciently low transaction costs some statistically signi?cant risk-corrected out-of-sample forecasting power is found for the Asian, Latin American, Middle East and Russian stock market indices.

 

Table 5.18 summarizes for all transaction costs cases the results of testing the set of 787 trend following technical trading techniques on the DJIA and on stocks listed in the DJIA (Chapter 3), on the AEX-index and on stocks listed in the AEX-index (Chapter 4) and on 51 stock market indices world wide (Chapter 5). If the return of the best technical trading strategy, selected in sample, in excess of the risk free interest rate is regressed against a constant α and the return of a market portfolio in excess of the risk free interest rate (see CAPMs (3.5), (4.1) and (5.1)), then the rows labeled “(1) in-sample CAPM: α > 0” show for each chapter the fraction of data series for which the estimate of α is signi?cantly positive at the 10% signi?cance level. The rows labeled “(2) pW < 0.10” show the fraction of data series for which White’s (2000) RC p-value is smaller than 0.10. The rows labeled “(3) pH < 0.10” show the fraction of data series for which Hansen’s (2001) SPA-test p-value is smaller than 0.10. Finally, the rows labeled “(4) out-of-sample CAPM: α > 0” show as in the rows labeled “(1) in-sample CAPM: α > 0” the fraction of data series for which the estimate of α is signi?cantly positive at the 10% signi?cance level, but this time when the return of the best recursive optimizing and training procedure in excess of the risk free interest rate is regressed against a constant α and the return of a market portfolio in excess of the risk free interest rate. Panel A shows the results if the best technical trading strategy is selected by the mean return criterion and panel B shows the results if the best technical trading strategy is selected by the Sharpe ratio criterion. In each chapter for all data series a technical trading strategy that is capable of beating the buy-and-hold benchmark can be selected in sample. In the case of zero transaction costs it can be seen in the rows labeled “(1) in-sample CAPM: α > 0” that in each chapter for a majority of the data series the estimate of α is signi?cantly positive, indicating that the best selected technical trading rule has statistically signi?cant forecasting power after.

 

Correction for risk. If transaction costs increase, then the number of data series for which a signi?cantly positive estimate of α is found declines. This can especially be observed in the results for the US stock market in Chapter 3 for which the fraction of data series for which a signi?cantly positive estimate of α is found declines to one quarter if 1% transaction costs are implemented. However, in the case of 1% transaction costs, for approximately half of the Dutch stock market data in Chapter 4 and for approximately half of the stock market indices in Chapter 5, the estimate of α is still signi?cantly positive. If the in-sample CAPM estimation results are compared with the out-of-sample CAPM estimation results, then the results in favour of technical trading of the latter tests are obviously worse than the results of the former tests. However, if transaction costs are zero, then in each chapter a group of data series can be found for which technical trading shows signi?cant out-of-sample forecasting power, after correction for risk. As transaction costs increase, this group becomes smaller and smaller. White’s (2000) RC and Hansen’s (2001) SPA-test are utilized to correct for data snooping. If little costs are implemented, then for the US stock market data in Chapter 3, the RC does not reject the null of no superior forecasting ability of the best selected technical trading rule over the buy-and-hold benchmark for all data series for both selection criteria. For the Dutch stock market data in Chapter 4 the same conclusion can be made, although the results in favour of technical trading are stronger, if the Sharpe ratio criterion is used. For a group of stock market indices in Chapter 5, in the case of zero transaction costs, it is found that the null hypothesis of no superior forecasting ability is rejected, especially if the Sharpe ratio criterion is used. However, if transaction costs increase to 1%, then for almost all data series the null hypothesis is not rejected anymore. The SPA-test corrects for the inclusion of poor and irrelevant strategies. Di?erences between the RC and SPA-test can especially be seen in Chapters 4 and 5, if the Sharpe ratio selection criterion is used. Then, for both the Dutch stock market data and the local main stock market indices, if 1% transaction costs are implemented, for more than one quarter of the data series the null hypothesis of no superior forecasting ability is rejected. Thus the biased RC leads in numerous cases to the wrong inferences. If no transaction costs are implemented, then technical trading shows economically and statistically signi?cant forecasting power for a group of data series, in all three chapters. In that case, generally, the results of the Sharpe ratio selection criterion are slightly better than the results of the mean return selection criterion. However, if transaction costs increase, then in Chapters 4 and 5 the Sharpe ratio selection criterion performs better in selecting the best technical trading strategy. If the Sharpe ratio criterion is used in selecting the best strategy, then for transaction costs up to 0.25%, technical trading.

 

Shows economically and statistically signi?cant forecasting power for approximately one fourth of the Dutch stock market data in Chapter 4. This is the case for approximately one third of the local main stock market indices in Chapter 5, if 0.50% transaction costs are implemented. It can be concluded that the DJIA and stocks listed in the DJIA are weak-form e?cient. That is, these data series are not predictable from their own price history at normal transaction costs. The AEX-index and stocks listed in the AEX-index are weak-form e?cient, only for transaction costs above 0.25%. For transaction costs below 0.25% pro?t opportunities exist. From the results in Chapter 5 it can be concluded that technical analysis applied to the stock market indices of emerging markets in Asia, Latin America, the Middle East and Russia has statistically signi?cant forecasting power only for low transaction costs (≤ 0.25% per trade), while for the Japanese, Northern American and Western European stock market indices the null hypothesis of weak-form e?ciency cannot be rejected for all transaction costs cases.

 

No transaction costs are implemented, for each index listed in the ?rst column. Column 2 shows the strategy parameters. Columns 3 and 4 show the mean return and excess mean return on a yearly basis in %/100 terms. Columns 5 and 6 show the Sharpe and excess Sharpe ratio. Column 7 shows the largest cumulative loss of the strategy in %/100 terms. Columns 8, 9 and 10 show the number of trades, the percentage of pro?table trades and the percentage of days these pro?table trades lasted. The last column shows the standard deviation of returns during pro?table trades divided by the standard deviation of returns during non-pro?table trades. The results are computed for an US-based trader who applies the technical trading rule set to the local main stock market indices recomputed in US Dollars. The daily interest rate on 1-month US certi?cates of deposits is used to compute the Sharpe and excess Sharpe ratio in columns 5 and 6.

 

Chapters 2 through 5 of this thesis contain empirical analyses whether technical trading has statistically signiÞcant forecasting power and yields economically signiÞcant proÞts when applied to Þnancial time series. The present chapter builds a simple theoretical Þnancial market model with fundamentalists and technical analysts. An important question in heterogeneous agents modeling is whether irrational traders can survive in the market, or whether they would lose money and are driven out of the market by rational investors, who would trade against them and drive prices back to fundamentals, as argued by e.g. Friedman (1953). In the last decade a number of theoretical and/or computational heterogeneous agent models, with fundamentalist traders competing against technical analysts, have been developed, see e.g. in Frankel and Froot (1988), De Long et al. (1989, 1990), Kirman (1991), Wang (1994), Lux (1995), Arthur et al. (1997), Brock and Hommes (1997, 1998), Farmer (1998), Hong and Stein (1999) and LeBaron et al. (1999). A common feature of these contributions is that technical traders may at times earn positive proÞts, survive evolutionary competition and need not be driven out of the market by trading strategies based upon economic fundamentals. Brock and Hommes (1998) investigate the dynamical behavior of a simple Þnancial market model with heterogeneous adaptively learning traders, where the fraction of traders following a certain forecasting rule changes over time.

 

To choose from a Þnite set of fundamental and trend following trading techniques. How many traders are using a particular technique in predicting prices depends on the past performances of these techniques, as measured by past proÞts or forecasting accuracy. Emphasis is placed on the change in dynamical behavior when the intensity of choice parameter, measuring how quickly agents switch between forecasting techniques, is varied. It is found that increasing this intensity of choice can lead to market instability and the emergence of complicated dynamics for asset prices and returns, with irregular switching between phases where prices are near to the fundamental value and phases of optimism where traders extrapolate trends. An extremely rich asset price dynamics emerges, with bifurcation routes to strange attractors, especially if switching to more successful strategies becomes more rapid. It is also found that even when costs of information gathering and trading are zero, then fundamentalists are in general not able to drive other trader types out of the market. Thus it is concluded that simple technical trading rules may survive evolutionary competition in a heterogeneous world where prices and beliefs coevolve over time and that therefore the Friedman argument should be considered with care. See e.g. Hommes (2001) for a survey and an extensive discussion of these points. One of the goals of heterogeneous agents modeling is to develop simple Þnancial asset pricing models that mimic the well-known characteristics of real Þnancial return distributions, such as little autocorrelation in the returns, volatility clustering and fat tails. Gaunersdorfer and Hommes (2000) develop a model in which volatility clustering becomes an endogenous phenomenon by the interaction of heterogeneous agents. Volatility clustering is caused by the coexistence of attractors, a stable fundamental steady state and a stable (quasi) periodic cycle. The time series properties of the model are compared with the daily closing prices of the S&P 500 in the period August 1961 through May 2000 and furthermore a GARCH model is estimated. It is concluded that the model approximates reality fairly well. This chapter is an extension of the Brock and Hommes (1998) model in that it adds a real moving-average technical trading strategy to the set of beliefs the traders can choose from. Moving averages are well known and frequently used prediction rules in Þnancial practice. They are intended to smooth out an otherwise volatile time series and to show its underlying directional trend. Furthermore, the model proposed in this chapter assumes that traders have constant relative risk aversion. That is, every trader in a given belief group invests the same proportion of his individual wealth in the risky asset. Hence, traders take the same amount of risk relative to their wealth. In the Brock and Hommes (1998) model, in contrast, it is assumed that the traders have constant absolute risk aversion. Irrespective of their individual wealth every trader in a certain belief group will.

 

Buy or sell short the same amount of stocks. Thus traders with less wealth are prepared to take greater relative risks than traders with more wealth. It should be noted that in the case of zero supply of outside stocks, the model developed in this chapter reduces to the Brock and Hommes (1998) model. In nonlinear dynamical models it is in general impossible to obtain explicit analytic expressions for the periodic and chaotic solutions. Therefore in applied nonlinear dynamics it is common practice to use a mixture of theoretical and numerical methods to analyze the dynamics. We perform a bifurcation analysis of the steady state by using numerical tools, such as delay and phase diagrams, bifurcation diagrams and the computation of Lyapunov exponents. In particular we show analytically that the fundamental steady state may become unstable due to a Hopf bifurcation. In section 6.2 the Brock and Hommes (1998) Þnancial market model with adaptively learning agents is reviewed. Thereafter, in section 6.3, the heterogeneous agents model with fundamentalists versus moving average traders, resulting in an eight dimensional nonlinear dynamical system, is derived. In section 6.4 a procedure is developed to determine trading volume. Section 6.5 presents an analytical stability analysis of the fundamental steady state. The eigenvalues of the linearized system are computed and it is examined which kind of bifurcations can occur. In section 6.6 numerical simulations are used to study the dynamical behavior of the model, especially when the steady state is locally unstable. Finally section 6.7 summarizes and concludes.

 

Bifurcations We have seen in equation (6.45) that in the case of risk neutrality of the fundamental traders, i.e. a = 0, there is locally always convergence to the fundamental steady state. If a = 4, then for β = 0 the dynamical system exhibits quasi periodic behavior and no change in the dynamics occurs by increasing β . Only for a < 0.456 changes in the dynamical behavior can be observed by varying β . Therefore we set the risk aversion parameter a initially low (0.42), so that the local dynamics around the steady state is dependent on the intensity of choice parameter β . For a = 0.42 Þgure 6.3a shows the bifurcation diagram with respect to β . A Hopf bifurcation occurs at β H = 635. Figure 6.3b shows the corresponding largest LCE plot. Before the Hopf bifurcation occurs the largest LCE is clearly smaller than zero, indicating convergence to the steady state. After the Hopf bifurcation occurred, the largest LCE is close to zero, indicating quasi periodic dynamical behavior. Thus for costs and low risk aversion for the fundamental traders and low intensity of choice for all traders, the price locally converges to the fundamental value. However for high intensity of choice, traders quickly change to the most proÞtable strategy and the moving-average trading strategy can survive in the market even for low risk aversion of the fundamental traders. Price ßuctuations are then driven by the evolutionary dynamics between the two di?erent beliefs. If the costs for the fundamental traders decrease to zero, then locally when varying β there is always convergence to the fundamental steady state, for low risk aversion. Fundamental expectations then dominate the moving-average strategy. Hence, costs can cause the fundamental steady state to become unstable, even if the risk aversion of fundamental traders is low. In the case of no costs and β = 250, Þgure 6.4a shows the bifurcation diagram with respect to the parameter a, when a is varied between 0.1 and 5. Figure 6.4b shows the corresponding largest LCE plot. At aH = 0.456 a Hopf bifurcation occurs and the dynamics shows quasi periodic behavior after the Hopf bifurcation. 

 

Mental traders become more risk averse, then even in the no cost case, moving average traders can survive in the market and a?ect the price by their actions. We set a equal to 4 and study the local dynamical behavior when varying the exponential moving average parameter µ. To observe a change in the dynamical behavior for the parameter µ we double the parameter λ to 14 basis points and we decrease the intensity of choice parameter β to 125. Figure 6.5a shows the bifurcation diagram with respect to µ, if µ is varied between 0.04 and 0.98. Figure 6.5b shows the corresponding largest LCE plot. Remember that by increasing the parameter µ the moving average follows the price series more closely and generates earlier a trading signal when the directional trend in prices changes direction. From the bifurcation diagram and the LCE plot it can be seen that the fundamental steady state becomes locally stable if the technical traders use a very fast moving average (µ > 0.82), that is if the technical traders quickly change their trading position if the directional trend in prices changes direction. For lower values of µ the LCE plot is close to zero and thus the dynamical system exhibits quasi periodic behavior. Price simulations Figures 6.6a, b, c and d show, given the parameter values in section 6.6.2, the time series plots of the price, return, fraction of fundamental traders and trading volume. The price series plot shows that there is a slow movement away from the fundamental value and a quick movement back. In Þgure 6.6a price starts below the fundamental value of 1000 and slowly increases with a declining positive return, or stated di?erently, the price sequence is concave. As price is increasing further and further above the fundamental value of 1000, the fundamental traders go short a larger fraction of their wealth, causing volume to increase as can be seen in Þgure 6.6d. The fraction of fundamental traders starts below 0.50 and is slowly increasing until the point that stock returns become smaller than the risk-free interest rate. Then the moving average forecasting rule is not proÞtable anymore and the fraction of fundamental traders increases sharply until approximately 0.56. These fundamental traders cause the price to turn back in the direction of the fundamental value. This change in trend is picked up by the moving average traders and they reinforce the downtrend by holding also short positions in the risky asset. Because as well the fundamental traders as the moving average traders are expecting price to decline, price falls quickly back to the fundamental value in a convex way. However, because the moving average traders are doing better than the fundamental traders, the fraction of fundamental traders declines sharply.

 

Majority of the agents was following the fundamental forecasting rule and already had short positions before the turn in price direction, volume drops sharply after the change of direction in the price trend. As price returns to the fundamental value, agents following the fundamental belief are closing their short positions, while traders following the moving average belief are holding more and more short positions, causing volume to increase. After prices dropped back to the fundamental value, prices keep on declining due to the moving average traders, with negative but increasing returns, so that the price sequence is convex. Volume increases, because traders following the fundamental belief are now holding more and more long positions as price moves below the fundamental value. Then, if the short position held by the moving average traders is not proÞtable anymore, the fraction of fundamental traders increases sharply turning the downward trend in price to an upward trend in price. The moving average traders detect the change in trend and will change their short position to a long position in the risky asset, causing price to increase back to the fundamental value. Because a majority of the agents was following the fundamental forecasting rule and already had long positions, volume drops sharply after the change of direction in the price trend. The price cycle is thus characterized by a period of small price changes when moving average traders dominate the market and periods of rapid decrease or increase of prices when fundamental traders temporarily dominate the market. Furthermore, volume goes by the prevailing trend as can be seen in Þgure 6.6d. That is, if the primary trend is upwards, then volume increases. Volume drops during a change in directional trend. Then, if the primary trend is downwards, volume also increases. This is a very important concept in technical analysis and the relation has been shown in many price charts. Adding dynamic noise to the deterministic skeleton leads to irregular price behavior as can be seen in Þgure 6.7a. Clearly periods with trending behavior can be identiÞed. Figure 6.7c shows that the fraction of fundamental traders is switching irregular between its lower- and upperbound. Because little autocorrelation, volatility clustering and fat tails are important characteristics of real Þnancial time series, we check our return series for these features. Figures 6.8b and 6.8c show the autocorrelation function plots of the returns and the squared returns up to order 36. Figure 6.8b shows that the return series does not exhibit any serial autocorrelation, which means that price changes are linearly independent. Further, according to Þgure 6.8c the squared return series does not exhibit any serial autocorrelation, which means that there is no volatility clustering present in the data. The return distribution does show excess kurtosis relatively to the normal distribution (see Þgure 6.8a). Thus our theoretical heterogeneous agents model only fails in mimicking the feature of volatility clustering.

 

In this chapter we have built a Þnancial market model with heterogeneous adaptively learning agents, fundamentalists and technical traders. The model is an extension of the Brock and Hommes (1998) model in that it extends the set of trading techniques the agents can independently choose from with a realistic moving-average technical trading rule. Moving averages are well known and one of the mostly used technical indicators in Þnancial practice and therefore they deserve to be implemented in heterogeneous agents modeling. Furthermore, the model is derived under the assumption of relative risk aversion, instead of absolute risk aversion as in the Brock and Hommes (1998) case. The model is derived under the assumption of inÞnitely many agents, who only di?er in the forecasting rule they select each period. Under the assumption that each agent has zero market power at each date, that is his individual investment decision will not inßuence the equilibrium price, it is shown that the fraction of total market wealth invested by all agents according to a certain belief converges in probability to the probability that the belief is chosen by the agents. Under the assumption of zero supply of outside stocks and the use of certain beliefs types it turns out that the price equilibrium formula is exactly the same as in Brock and Hommes (1998), namely that the price is equal to the discounted value of the average expected price and dividends by all agents. Moreover if the moving-average technical trading rule is added to the model, then also risk aversion and expected dispersion of future returns play a role in our model. In the end, our Þnancial market model is an eight dimensional nonlinear dynamical system. The steady state price is equal to the fundamental value, which is the discounted value of all future dividends. Analytically we derive the eigenvalues of the linearized system and we examine for which parameter values bifurcations occur. It is shown that the system only can exhibit a Hopf bifurcation. We use numerical tools such as delay, phase and bifurcation diagrams, and computation of Lyapunov characteristic exponents to study the local stability around the fundamental steady state. If there is no di?erence in costs of applying the fundamental or moving-average strategy, then it is found that the intensity of choice parameter, measuring how quickly traders switch beliefs, has no inßuence on the dynamical behavior. In the presence of costs, if the risk aversion parameter of the fundamental traders is low enough, then these traders always drive prices back to the fundamental steady state for the case the intensity of choice parameter is su?ciently low. For high values of the intensity of choice parameter, even for low risk aversion, quasi periodic price behavior can occur as a consequence of a Hopf bifurcation. If costs of all trader types are set to zero and if more realistic values for the risk aversion parameter are.

 

Chosen, then fundamental traders are too risk averse to drive prices to the fundamental steady state and the price exhibits quasi periodic behavior. However, if the risk aversion parameter is high and the technical traders use a very fast moving average, which follows the price closely, then the price does converge to the fundamental value. We study a case in which we choose parameter values that are economically sensible. The solution of the dynamical system is quasi periodic price behavior. Interaction between fundamentalists and technical analysts may thus destabilize the market and lead to persistent price ßuctuations around an unstable fundamental steady state. It turns out that fundamental traders change the direction of a prevailing price trend, but that once the direction has changed, the technical traders push prices back to the fundamental value. Moreover it is found that volume goes by the prevailing trend, that is if the primary trend is upwards or downwards, then volume increases, only dropping if a change in the direction of the trend occurs. This is an important concept in technical analysis. Dynamic noise to the deterministic skeleton is added and leads to irregular price behavior. The features of the return distribution of the dynamical system are examined, but it is concluded that although the model generates returns series which show zero autocorrelation and fat tails, the model fails in mimicking the important characteristic of volatility clustering.

 

References

Achelis, S.B. (1995), Technical Analysis from A to Z, McGraw-Hill, New York. Adriaanse, G. (2002), Technical Indicators, master thesis, University of Amsterdam. Alexander, S.S. (1961), Price Movements in Speculative Markets: Trends or Random Walks, Industrial Management Review 2, 7-26. Alexander, S.S. (1964), Price movements in speculative markets: Trends or random walks, Number 2, Industrial Management Review 5, 25-46. Allen, H., Taylor, M.P. (1990), Charts, Noise and Fundamentalists in the London Foreign Exchange Market, The Economic Journal 100, 49-59. Arditti, F.D., McCollough, W.A. (1978), Can Analysts Distinguish Between Real and Randomly Generated Stock Prices? Financial Analysts Journal, 70-74. Arthur, W.B., Holland, J.H., LeBaron, B., Palmer, R., Taylor, P. (1997), Asset pricing under endogenous expectations in an artiÞcial stock market. In: Arthur, W.B., Durlauf, S.N., and Lane, D.A., eds., The economy as an evolving complex system II, Redwood City, Addison-Wesley. Bartlett, M.S. (1946), On the Theoretical SpeciÞcation of Sampling properties of Autocorrelated Time Series, Journal of the Royal Statistical Society, Series B, 8, 27-41. Bass, A.B. (1999), The Predictors: How a band of maverick physicists set out to beat Wall Street, Penguin Books Ltd, London. Bera, A.K., Higgins, M.L. (1993), Arch Models: Properties, Estimation and Testing, Journal of Economic Surveys 7, 305-366. Bernstein, P.L. (1996), Against the Gods: the remarkable story of risk, John Wiley & Sons, Inc. Bessembinder, H., Chan, K. (1995), The proÞtability of technical trading strategies in the Asian stock markets, PaciÞc-Basin Finance Journal 3, 257-284. Bessembinder, H., Chan, K. (1998), Market E?ciency and the Returns to Technical Analysis, Financial Management 27, 5-17. Bodie, Z., Kane, A., Marcus, A.J. (1996), Investments, Irwin.

 

References

 

Bollerslev, T., Wooldridge, J.M. (1992), Quasi-Maximum Likelihood Estimation and Inference in Dynamic Models with Time Varying Covariances, Econometric Reviews 11, 143-172. Boswijk, H.P. (1998), Unit Roots and Cointegration, University of Amsterdam, Lecture notes. Bouman, S., Jacobsen, B. (1997), The Halloween Indicator: Sell in May and go away, Chapter in thesis: Time Series Properties of Stock Returns, 86-106, University of Amsterdam. Box, G., Pierce, D. (1970), Distribution of Autocorrelations in Autoregressive Moving Average Time Series Models, Journal of the American Statistical Association 65, 15091526. Brock, W.A., Hommes, C.H. (1997), A rational route to randomness, Econometrica 65, 1059-1095. Brock, W.A., Hommes, C.H. (1997), Models of complexity in economics and Þnance, System Dynamics in Economic and Financial Models, ed C. Hey et al (New York: Wiley), 3-41. Brock, W.A., Hommes, C.H. (1998), Heterogeneous beliefs and routes to chaos in a simple asset pricing model, Journal of Economic Dynamics and Control 22, 1235-1274. Brock, W.A., Lakonishok, J., LeBaron, B. (1992), Simple Technical Trading Rules and the Stochastic Properties of Stock Returns, The Journal of Finance 47, 1731-1764. Campbell J.Y., Lo, A.W., MacKinlay, A.C. (1997), The Econometrics of Financial Markets, Princeton University Press, New Jersey. Chan, L.K.C., Lakonishok, J. (1993), Institutional Trades and Intraday Stock Price Behavior, Journal of Financial Economics, 173-199. Cheung, Y., Chinn, M.D. (1999), Macroeconomic Implications of the Beliefs and Behavior of Foreign Exchange Traders, NBER working paper 7417. Cheung, Y., Chinn, M.D., Marsh, I.W. (2000), How Do UK-Based Foreign Exchange Dealers Think Their Market Operates? NBER working paper 7524. Cootner, P. (ed.) (1964), The Random Character of Stock Market Prices, MIT Press, Cambridge. Reprint in 2000 by Risk Publications, London. Coutts, J.A., Cheung, K.C. (2000), Trading rules and stock returns: some preliminary short run evidence from the Hang Seng 1985-1997, Applied Financial Economics 10, 579-586. Cowles, A. (1933), Can Stock Market Forecasters Forecast?, Econometrica 1, 309-324. Cowles, A. (1944), Stock Market Forecasting, Econometrica, 206-214.

 

References

 

Curcio, R., Goodhart, G. Guillaume, D., Payne, R. (1997), Do Technical Trading Rules Generate ProÞts? Conclusions from the Intra-Day Foreign Exchange Market, International Journal of Finance and Economics 2, 267-280. Cuthbertson, K. (1996), Quantitative Financial Economics: Stocks, Bonds and Foreign Exchange, John Wiley & Sons Ltd., London. Dacorogna, M.M., M¨ uller, U.A., Pictet, O.V. (1991), A Measure of Trading Model Performance with a Risk Component, A discussion paper by the O&A Research Group, MMD.1991-05-24. Dash, M. (1999), Tulipomania: The Story of the World’s Most Coveted Flower and the Extraordinary Passions It Aroused, Victor Gollancz, London. De Long, J.B., Shleifer, A., Summers, L.H., Waldmann, R.J. (1989), The size and incidence of the losses of noise trading, Journal of Finance 44, 681-696. De Long, J.B., Shleifer, A., Summers, L.H., Waldmann, R.J. (1990), Noise trader risk in Þnancial markets, Journal of Political Economy 98, 703-738. Detry, P.J., Gregoire, P. (2001), Other Evidences of the Predictive Power of Technical Analysis: the Moving-Average Rules on European Indices, working paper, European Financial Management Association. Diebold, F.X. (1986), Testing for Serial Correlation in the Presence of ARCH, Proceedings of the American Statistical Association, Business and Economic Statistics Section, 323328. Diebold, F.X., Mariano, R.S. (1995), Comparing predictive accuracy, Journal of Business and Economic Statistics 13, 253-265. Dooley, M.P., Shafer, J.R. (1976), Analysis of Short-Run Exchange Rate Behavior: March 1973 to September 1975, International Finance Discussion Paper 123, Federal Reserve Board, Washington, D.C. Dooley, M.P., Shafer, J.R. (1983), Analysis of Short-Run Exchange Rate Behavior: March 1973 to November 1981, 43-69, Exchange Rate and Trade Instability: Causes, Consequences and Remedies, in D. Bigman and T.Taya eds., Ballinger, Cambridge, MA. Dudewicz, E.J., Mishra, S.N. (1988), Modern Mathematical Statistics, John Wiley & Sons, Inc. Edwards, R.D., Magee, J. (1998), Technical Analysis of Stock Trends. Seventh Edition, second printing, John Magee, Inc. Efron, B. (1979), Bootstrap methods: Another look at the jackknife, The Annals of Statistics 7, 1-26. Efron, B., Tibshirani, R. (1986), Bootstrap methods for standard errors, conÞdence intervals and other measures of statistical accuracy, Statistical Science 1, 54-77.

 

References

 

Enders, W. (1995), Applied Econometric Time Series, John Wiley & Sons, Inc. Fama, E.F. (1965a), The behavior of stock market prices, Journal of Business 38, 34-105. Fama, E.F. (1965b), Tomorrow on the New York Stock Exchange, Journal of Business 38, 285-299. Fama, E.F. (1970), E?cient Capital Markets: A Review of Theory and Empirical Work, Journal of Finance 25, 383-417. Fama, E.F., Blume, M.E. (1966), Filter Rules and Stock-Market Trading, Journal of Business 39, 226-241. Fama, E.F, French, K.R (1988), Dividend yields and expected stock returns, Journal of Financial Economics 22, 3-25. Farmer, J.D. (1998), Market force, ecology, and evolution, Santa Fe Institute working paper 98-12-117. Fern´ andez-Rodr´ õguez, F., Sosvilla-Rivero, S., Andrada-F´ elix, J. (2001), Technical Analysis in the Madrid Stock Exchange, Moneda y Cr´ edito 213, 11-37. Frankel, J.A., Froot, K.A. (1988), Chartists, fundamentalists and the demand for dollars, Greek Economic Review 10, 49-102. Franses, P.H., Dijk, D. van (2000), Non-linear time series models in empirical Þnance, Cambridge University Press, New York. Freedman, D. (1984), On bootstrapping two-stage least squares estimates in stationary linear models, Annals of Statistics 12, 827-842. Freedman, D., Peters, S. (1984a), Bootstrapping a regression equation: Some empirical results, Journal of the American Statistical Society 79, 97-106. Freedman, D., Peters, S. (1984b), Bootstrapping an econometric model: Some empirical results, Journal of Business and Economic Statistics 2, 150-158. Friedman, M. (1953), The case of ßexible exchange rates, In: Essays in Positive Economics, University of Chicago Press, Chicago. Friend, I., Blume, M.E. (1975), The Demand for Risky Assets, The American Economic Review 65, 900-922. Gaunersdorfer, A. (2000), Endogenous ßuctuations in a simple asset pricing model with heterogeneous agents, Journal of Economic Dynamics and Control 24, 799-831. Gaunersdorfer, A., Hommes, C.H. (2000), A nonlinear structural model for volatility clustering, CeNDEF Working Paper 00-02, University of Amsterdam. Gaunersdorfer, A., Hommes, C.H., Wagener, F.O.O. (2000), Bifurcation Routes to Volatility Clustering, CeNDEF Working Paper 00-09, University of Amsterdam. Giersbergen, N.P.A. van (1998), Bootstrapping Dynamic Econometric Models, Thesis, University of Amsterdam.

 

References

 

Graham, B. (1949), The Intelligent Investor, Fourth revised edition (1973), Harper & Row, Publishers, Inc., New York. Grandia, V. (2002), The Search for the Golden Rule: A Reality Check of Technical Analysis in the Dutch Stock Market, master thesis, University of Amsterdam. Grossman, S.J., Shiller, R.J. (1981), The Determinants of the Variability of Stock Market Prices, The American Economic Review, 222-227. Hamilton, W.P. (1922), The Stock Market Barometer, Harper & Brothers, New York. Reprint in 1998 by John Wiley & Sons, Inc., New York. Hansen, P.R. (2001), An Unbiased and Powerful Test for Superior Predictive Ability, Brown University, Department of Economics, working paper no.01-06. Hommes, C.H. (1997), Nonlinear Economic Dynamics, University of Amsterdam, Lecture notes. Hommes, C.H. (2001), Financial markets as nonlinear adaptive evolutionary systems, Quantitative Finance 1, 149-167. Hong, H., Stein, J.C. (1999), A uniÞed theory of underreaction, momentum trading, and overreaction in asset markets, Journal of Finance 54, 2143-2184. Hsieh, D.A. (1988), The Statistical Properties of Daily Foreign Exchange Rates: 19741983, Journal of International Economics 24, 129-145. Hudson, R., Dempsey, M., Keasey, K. (1996), A note on the weak form e?ciency of capital markets: The application of simple technical trading rules to UK stock prices - 1935 to 1994, Journal of Banking and Finance 20, 1121-1132. Hull, J.C. (1991), Futures and Options Markets, Prentice-Hall, Inc., New Jersey. Isakov D., Hollistein, M. (1999), Application of simple technical trading rules to Swiss stock prices: Is it proÞtable?, Financial Markets and Portfolio Management 13, 9-26. James, F.E. (1968), Monthly Moving Averages: An E?ective Investment Tool?, Journal of Financial and Quantitative Analysis, 315-326. Jensen, M.C. (1967), Random Walks: Reality or Myth - Comment, Financial Analysts Journal, 77-85. Jensen, M.C., Benington, G.A. (1969), Random Walks and Technical Theories: Some Additional Evidence, Journal of Finance 25, 469-482. Kendall, M.G. (1953), The Analysis of Economic Time-Series, Journal of the Royal Statistical Society 96, 11-25. Keynes, J.M. (1936), The General Theory of Unemployment, Interest and Money, Harcourt Brace, London.

 

References

 

Kho, B.C. (1996), Time-varying risk premia, volatility, and technical trading rule proÞts: Evidence from foreign currency futures markets, Journal of Financial Economics 41, 249-290. Kirman, A. (1991), Epidemics of opinion and speculative bubbles in Þnancial markets, In: M. Taylor (ed.), Money and Þnancial markets, Macmillan, London. Knez, P., Ready, M. (1996), Estimating the ProÞts from Trading Strategies, Review of Financial Studies, 1121-1164. Langedijk, N. (2001), The Predictability of Technical Trading Rules on the Foreign Exchange Market: A Bootstrap Approach, master thesis, University of Amsterdam. LeBaron, B. (1993), Practical Comparisons of Foreign Exchange Forecasts, Neural Network World 6, 779-790. LeBaron, B. (2000a), The Stability of Moving-Average Technical Trading Rules on the Dow-Jones Index, Derivatives Use, Trading and Regulation 5, 324-338. LeBaron, B. (2000b), Technical Trading ProÞtability in Foreign Exchange Market’s in the 1990’s, Brandeis University, working paper. LeBaron, B., Arthur, W.B., Palmer, R. (1999), Time series properties of an artiÞcial stock market, Journal of Economic Dynamics and Control 23, 1487-1516. Lee, C.I., Mathur, I. (1995), Trading rule proÞts in European currency spot cross rates, Journal of Banking and Finance 20, 949-962. Leeson, N. (1996), Rogue Trader, Time Warner Paperbacks. Levich, R.M., Thomas L.R. (1993), The signiÞcance of technical trading-rule proÞts in the foreign exchange market: A bootstrap approach, Journal of International Money and Finance 12, 451-474. Levy, R.A. (1967), Relative Strength as a Criterion for Investment Selection, Journal of Finance 22, 595-610. Levy, R.A. (1971), The Predictive SigniÞcance of Five-Point Chart Patterns, The Journal of Business 44, 316-323. Lintner, J. (1965), The Valuation of Risk Assets and the Selection of Risky Investments in Stock Portfolios and Capital Budgets, Review of Economics and Statistics 47, 13-37. Ljung, G., Box, G. (1978), On a Measure of Lack of Fit in Time Series Models, Biometrika 65, 297-303. Lo, A.W., MacKinlay, A.C. (1988), Stock market prices do not follow random walks: evidence from a simple speciÞcation test, Review of Financial Studies 1, 41-66. Lo, A.W., MacKinlay, A.C., (1997), Maximizing predictability in the stock and bond markets, Macroeconomic Dynamics 1, 102-134.

 

References

 

Lo, A.W., MacKinlay, A.C. (1999), A non-random walk down Wall Street, Princeton University Press, Princeton. Lo, A.W., Mamaysky, H., Wang, J. (2000), Foundations of technical analysis: computational algorithms, statistical inference and empirical implementation, Journal of Finance 55, 1705-1722. Lux, T. (1995) Herd Behavior, Bubbles and Crashes, The Economic Journal 105, 881-896. Malkiel, B.G. (1996), A Random Walk down Wall Street, W.W. Norton & Company, Inc, New York. Mandelbrot, B. (1963), The Variation of Certain Speculative Prices, Journal of Business 36, 394-419. Markowitz, H.M. (1952), Portfolio Selection, Journal of Finance 7, 77-91. Markowitz, H.M. (1959), Portfolio Selection: E?cient DiversiÞcation of Investments, John Wiley, New York. Marquering, W., Verbeek, M. (2000), The Economic Value of Predicting Stock Index Returns and Volatility, Tilburg University, Center for Economic Research, Discussion paper 78. Menkho?, L. (1998), The noise trading approach - questionnaire evidence from foreign exchange, Journal of International Money and Finance 17, 547-564. Mills, T.C. (1990), Time series techniques for economists, Cambridge University Press. Mills, T.C. (1997), Technical Analysis and the London Stock Exchange: Testing Trading Rules Using the FT30, International Journal of Finance and Economics 2, 319-331. Ming Ming, L., Mat Nor, F., Krishnan Guru, B. (2000), Technical Analysis in the Malaysian Stock Market: An Empirical Evidence, working paper, Universiti Kebangsaan Malaysia. Murphy J.J. (1986), Technical Analysis of the Futures Markets, New York institute of Þnance. Neftci, S.N. (1991), Naive Trading Rules in Financial Markets and Wiener-Kolmogorov Prediction Theory: A Study of “Technical Analysis”, Journal of Business 64, 549-571. Nelson, D.B. (1995), Conditional heteroskedasticity in asset returns: a new approach, Econometrica 59, 347-370. Nelson, S.A. (1903), The ABC of Stock Speculation. Reprint in 1999 by Fraser Publishing Company, Burlington. Newey, W., West, K. (1987), A Simple Positive Semi-DeÞnite, Heteroskedasticity and Autocorrelation Consistent Covariance Matrix, Econometrica 55, 703-708. O’Hara, M. (1995), Market Microstructure Theory, Blackwell Publishers, Inc., Massachusettes.

 

References

 

Osler, C.L., Chang, P.H.K. (1995), Head and Shoulders: Not Just a Flaky Pattern, Sta? Reports No. 4, Federal Reserve Bank of New York. Pesaran, M.H., Timmermann, A. (1995), Predictability of Stock Returns: Robustness and Economic SigniÞcance, The Journal of Finance 50, 1201-1228. Pesaran, M.H., Timmermann, A. (2000), A Recursive Modeling Approach to Predicting UK Stock Returns, Economic Journal 110, 159-191. Phillips, P.C.B. (1986), Understanding Spurious Regressions in Econometrics, Journal of Econometrics 33, 311-340. Politis, D., Romano, J. (1994), The Stationary Bootstrap, Journal of the American Statistical Association 89, 1303-1313. Prestbo, J.A. (1999), The Market’s Measure, Dow Jones & Company, Inc., New York. Pring, M. (1998), Introduction to Technical Analysis, McGraw-Hill, New York. Ratner, M., Leal, R.P.C. (1999), Tests of technical trading strategies in the emerging equity markets of Latin America and Asia, Journal of Banking and Finance 23, 18871905. Ready, M.J. (1997), ProÞts from Technical Trading Rules, working paper, University of Wisconsin-Madison. Rhea, R. (1932), The Dow Theory, Barron’s, New York. Reprint in 1993 by Fraser Publishing Company, Burlington. Roberts, H.V. (1959), Stock-Market “Patterns” and Financial Analysis: Methodological Suggestions, Journal of Finance 14, 1-10. Samuelson, P. (1965), Proof that Properly Anticipated Prices Fluctuate Randomly, Industrial Management Review 6, 41-49. Satterthwhaite, R.E. (1946), Biom. Bull. 2 :110. Schulmeister, S. (1988), Currency Speculation and Dollar Fluctuations, Quarterly Review Banca Nazionale del Lavoro, 343-365. Sharpe, W. (1964), Capital Asset Prices: A Theory of Market Equilibrium, Journal of Finance 19, 425-442. Shiller, R.J. (2000), Irrational Exuberance, Princeton University Press, New Jersey. Snedecor, G.W., Cochran, W.G. (1989), Statistical Methods (8), Iowa: Stek University Press, Ames. Sullivan, R., Timmermann, A., White, H. (1999), Data-snooping, Technical Trading Rule Performance, and the Bootstrap, Journal of Finance 54, 1647-1691. Sullivan, R., Timmermann A., White, H. (2001), Dangers of data mining: The case of calendar e?ects in stock returns, Journal of Econometrics 105, 249-286.

 

References

 

Sweeney, R.J. (1986), Beating the Foreign Exchange Market, The Journal of Finance 61, 163-182. Sweeney, R.J. (1988), Some New Filter Rule Tests: Methods and Results, Journal of Financial and Quantitative Analysis 23, 285-300. Taylor, M.P., Allen, H. (1992), The use of technical analysis in the foreign exchange market, Journal of International Money and Finance 11, 304-314. Theil, H., Leenders, C.T. (1965), Tomorrow on the Amsterdam Stock Exchange, Journal of Business 38, 277-284. Thomson, R. (1998), Apocalypse Roulette, Pan Macmillan, London. Wang, J. (1994), A model of competitive stock trading volume, Journal of Political Economy 102, 127-168. West, K.D. (1996), Asymptotic Inference about Predictive Ability, Econometrica 64, 10671084. Westerho?, F.H. (2002), Expectations driven distortions in the foreign exchange market, Journal of Economic Behavior and Organization 1502, 1-24. White, H. (1980), A Heteroskedasticity-Consistent Covariance Matrix and a Direct Test for Heteroskedasticity, Econometrica 48, 817-838. White, H. (2000), A Reality Check for Data Snooping, Econometrica 68, 1097-1126. Williams, J.B. (1938), The Theory of Investment Value, Harvard University Press. Working, H. (1934), A Random Di?erence Series for Use in the Analysis of Time Series, American Statistical Association Journal 29, 11-24.

 

Samenvatting (Summary in Dutch)

Financi¨ ele analisten gebruiken “fundamentele” en “technische” analyse om de toekomstige prijsontwikkeling van een Þnanci¨ ele waarde, zoals een aandeel, termijncontract, valuta etc., te kunnen voorspellen. Bij fundamentele analyse wordt er onderzoek gedaan naar allerlei economische factoren die de inkomsten van een Þnanci¨ ele waarde, zoals dividenden, kunnen be¨ õnvloeden. Deze economische factoren worden ook wel de fundamentele variabelen genoemd. De fundamentele variabelen meten macro-economische omstandigheden, zoals bijvoorbeeld de olieprijs, inßatie, rente, werkloosheid, etc., bedrijfstak speciÞeke omstandigheden, zoals bijvoorbeeld concurrentie, technologische veranderingen, vraag/aanbod, etc., en bedrijfsspeciÞeke omstandigheden, zoals bijvoorbeeld dividend, groei, inkomsten, rechtszaken, stakingen, etc. Op basis van alle verzamelde fundamentele informatie wordt de fundamentele of intrinsieke waarde berekend. Vervolgens wordt bepaald of de marktprijs van de Þnanci¨ ele waarde lager of hoger is dan de fundamentele waarde en wordt de Þnanci¨ ele waarde gekocht of verkocht. Technische analyse is de bestudering van koerspatronen aan de hand van graÞeken met als doel het voorspellen van de toekomstige koersontwikkeling. De ÞlosoÞe achter technische analyse is dat alle informatie geleidelijk wordt verwerkt in de prijs van een Þnanci¨ ele waarde. Hierdoor bewegen koersen zich voort in min of meer regelmatige patronen, die herhaaldelijk zijn waar te nemen in de koersgraÞeken. Technische analisten claimen dat zij die patronen kunnen herkennen en daarop winstgevend kunnen handelen. Samengevat kan dus gezegd worden dat de technische analist het e?ect van een prijsverandering op de toekomstige koersontwikkeling bestudeert, terwijl de fundamentele analist altijd op zoek is naar een economische oorzaak voor een prijsverandering. In de huidige wetenschappelijke literatuur over Þnanci¨ ele markten staat de e?ci¨ ente markthypothese (EMH) nog steeds centraal. Een Þnanci¨ ele markt heet zwak e?ci¨ ent als het onmogelijk is om een handelsstrategie te ontwikkelen die op basis van de koershistorie van een Þnanci¨ ele waarde de toekomstige koersontwikkeling van die Þnanci¨ ele waarde kan voorspellen. Een Þnanci¨ ele markt heet semi-stringent e?ci¨ ent als het onmogelijk is om een handelsstrategie te ontwikkelen die op basis van alle publieke informatie de toekomstige 295.

 

koersontwikkeling van een Þnanci¨ ele waarde kan voorspellen. Tenslotte heet een Þnanci¨ ele markt sterk e?ci¨ ent als het onmogelijk is om op basis van alle denkbare beschikbare informatie, dus ook insider informatie, de toekomstige koersontwikkeling van een Þnanci¨ ele waarde te voorspellen. Bovendien geldt voor elk van de drie e?ci¨ entie hypothesen dat de handelsstrategie niet continu een bovengemiddeld rendement kan opleveren als er wordt gecorrigeerd voor risico en transactiekosten. Semi-stringente e?ci¨ entie impliceert zwakke e?ci¨ entie en sterke e?ci¨ entie impliceert semi-stringente en zwakke e?ci¨ entie. Als de zwakke vorm van de EMH verworpen kan worden, dan kan ook de semi-stringente en de sterke vorm van de EMH verworpen worden. Een bovengemiddeld rendement van technische handelsstrategie¨ en is dus in strijd met de zwakke vorm van de EMH. In dit proefschrift wordt de zwakke vorm van de EMH getest door het toepassen van vele verschillende trend-volgende technische handelsstrategie¨ en op een groot aantal Þnanci¨ ele datareeksen. Na correctie voor transactiekosten, risico en de zoektocht naar de beste strategie zal statistisch getoetst worden of de voorspelbaarheid en de winsten gegenereerd door technische handelsregels echt zijn of slechts schijn. In hoofdstuk 2 wordt een verzameling van 5350 technische handelsstrategie¨ en toegepast op de koersen van cacao goederen termijncontracten verhandeld op de London International Financial Futures Exchange (LIFFE) en op de New York Co?ee, Sugar and Cocoa Exchange (CSCE) in de periode van januari 1983 tot en met juni 1997. Voor diezelfde periode wordt de verzameling van strategie¨ en ook toegepast op de Pond-Dollar wisselkoers. Als de verzameling van handelsstrategie¨ en wordt toegepast op de prijzen van de LIFFE cacao termijncontracten, dan wordt er gevonden dat 58% van de technische handelsregels een strikt positief bovenmatig gemiddeld rendement oplevert, zelfs als er een correctie wordt gemaakt voor transactiekosten. Bovendien laat een groot percentage van de technische strategie¨ en een statistische signiÞcante voorspellende kracht zien. Echter, als dezelfde strategieverzameling wordt toegepast op de prijzen van de CSCE cacao termijncontracten, dan worden er veel slechtere resultaten gevonden. Nu levert slechts 12% van de handelsstrategie¨ en een bovenmatig gemiddeld rendement op. Verder wordt er nauwelijks nog enige statistische signiÞcante voorspellende kracht gevonden. Bootstrap technieken onthullen dat de goede resultaten die gevonden zijn voor de LIFFE cacao termijncontracten niet verklaard kunnen worden door enkele populaire econometrische tijdreeksmodellen, zoals het random walk, het autoregressieve, en het GARCH model. Echter, de resultaten lijken wel verklaard te kunnen worden door een model met een structurele verandering in de trend. Het grote verschil in de gevonden resultaten voor de LIFFE en CSCE cacao termijncontracten kan worden toegeschreven aan het vraag/aanbod mechanisme in de cacaomarkt in combinatie met een toevallige invloed van de Pond-Dollar wisselkoers.

 

De trends in de cacaoreeksen vallen toevallig samen met de trends in de Pond-Dollar wisselkoers, waardoor de prijstrends in de LIFFE termijncontracten worden versterkt, maar de prijstrends in de CSCE cacao termijncontracten worden afgezwakt. Verder suggereert deze casestudie een verband tussen het succes of falen van technische handelsregels en de relatieve grootte van een trend en de beweeglijkheid van een Þnanci¨ ele tijdreeks. In de hoofdstukken 3, 4 en 5 wordt een verzameling van trend volgende technische handelsstrategie¨ en toegepast op de koersen van verscheidene aandelen en op de indices van internationale aandelenmarkten. Twee verschillende maatstaven worden gebruikt om het resultaat van een strategie te beoordelen, namelijk het gemiddelde rendement en de Sharpe ratio. In de berekeningen wordt er gecorrigeerd voor transactiekosten. Als technische handelsregels winstgevend blijken te zijn, dan kan het zijn dat die winsten de beloning zijn voor het dragen van risico. Daarom worden er Sharpe-Lintner capital asset pricing modellen (CAPMs) geschat om deze hypothese te toetsen. Als technische handelsregels een economische signiÞcante winst opleveren na correctie voor risico en transactiekosten, dan bestaat het gevaar dat dit het resultaat is van een te uitgebreide zoektocht naar de best strategie (“data snooping”). Daarom wordt er de nul hypothese getoetst of de beste technische handelsregel daadwerkelijk superieur is ten opzichte van een passieve strategie van eenmaal kopen en niet meer verkopen, nadat er een correctie is uitgevoerd voor de zoektocht naar de beste handelsregel. Om deze hypothese te toetsen wordt er gebruik gemaakt van twee recentelijk ontwikkelde toetsen, zoals White’s (2000) Reality Check (RC) en Hansen’s (2001) test voor Superior Predictive Ability (SPA). Tenslotte wordt er met een recursieve methode van optimaliseren en toepassen getest of technische handelsregels daadwerkelijk een out-of-sample voorspellende kracht hebben. Bijvoorbeeld, aan het begin van elke maand wordt de technische handelsregel geselecteerd die de beste resultaten opleverde in het afgelopen half jaar en vervolgens wordt die strategie gebruikt om handelssignalen te genereren gedurende die maand. In hoofdstuk 3 wordt een verzameling van 787 trend volgende technische handelsstrategie¨ en toegepast op de Dow-Jones Industrial Index en op alle aandelen genoteerd in de Dow-Jones Industrial Index in de periode van januari 1974 tot en met juni 2001. Omdat uit verschillende wetenschappelijke artikelen naar de voorspelbaarheid van speculatieve prijsreeksen is gebleken dat technische handelsregels een statistische signiÞcante voorspellende kracht vertonen tot het jaar 1987, maar niet in de periode daarna, wordt de steekproef opgedeeld in de twee subperioden 1973-1986 en 1987-2002. In alle perioden wordt er zowel voor het gemiddeld rendement als voor het Sharpe ratio selectiecriterium gevonden dat voor elke datareeks een technische handelsregel kan worden geselecteerd die in staat is de passieve strategie van eenmaal kopen en vasthouden te verslaan, ook.

 

Als er wordt gecorrigeerd voor transactiekosten. Bovendien wordt er, wanneer er geen transactiekosten worden opgevoerd, met behulp van het regresseren van Sharpe-Lintner CAPMs voor de meeste datareeksen gevonden dat technische handelsregels een statistisch signiÞcant bovengemiddeld rendement opleveren, zelfs na correctie voor risico. Echter, als de transactiekosten toenemen dan wordt de nul hypothese dat door technische handelsregels gegenereerde winsten een beloning zijn voor het dragen van risico, voor steeds meer datareeksen niet verworpen. Tevens wordt bij 0.25% transactiekosten voor vrijwel alle onderzochte datareeksen de nul hypothese dat de beste technische handelsstrategie niet superieur is ten opzichte van de strategie van eenmaal kopen en vasthouden, nadat een correctie is uitgevoerd voor de zoektocht naar die beste strategie, niet verworpen door de RC en de SPA-test. Tenslotte vertoont de recursieve methode van optimaliseren en toepassen van handelsregels geen voor risico gecorrigeerde out-of-sample voorspellende kracht van technische analyse. Er kan dus worden geconcludeerd dat trend-volgende technische handelsregels, na correctie voor transactiekosten, risico en de zoektocht naar de beste strategie, geen economische en statistische signiÞcante voorspellende kracht vertonen voor zowel de Dow-Jones Industrial Index als de aandelen genoteerd in de Dow-Jones Industrial Index. In hoofdstuk 4 wordt de strategieverzameling van hoofdstuk 3 toegepast op de AEXindex en op 50 aandelen genoteerd in de AEX-index in de periode van januari 1983 tot en met mei 2002. Voor zowel het gemiddeld rendement als het Sharpe ratio selectiecriterium wordt er gevonden dat voor elke datareeks een technische handelsstrategie kan worden geselecteerd die in staat is om de strategie van eenmaal kopen en vasthouden te verslaan, zelfs na correctie voor transactiekosten. Bovendien wordt er voor ongeveer de helft van de onderzochte datareeksen gevonden dat de beste strategie een statistische signiÞcante voorspellende kracht heeft, ook na correctie voor risico. Vervolgens wordt er een correctie gemaakt voor de zoektocht naar de beste technische handelsregel met behulp van de RC en de SPA-test. Als het gemiddeld rendement criterium wordt gebruikt om de beste strategie te selecteren, dan leiden beide toets procedures tot dezelfde conclusie als minstens 0.10% transactiekosten worden opgevoerd: de beste geselecteerde technische handelsregel is niet statistisch signiÞcant superieur aan de strategie van eenmaal kopen en vasthouden. Echter, als het Sharpe ratio criterium wordt toegepast, dan wordt voor ongeveer ´ e´ en derde van de aandelen de nul hypothese van geen superieure voorspellende kracht na correctie voor de zoektocht naar de beste strategie wel verworpen, zelfs als 1% transactiekosten worden opgevoerd. In tegenstelling tot de resultaten gevonden in hoofdstuk 3 vinden we in hoofdstuk 4 dat technische analyse toekomstige koersontwikkelingen kan voorspellen, na correctie voor transactiekosten, risico en data snooping, als het Sharpe.

 

Ratio criterium wordt gebruikt om de beste strategie te selecteren. Tenslotte toont de recursieve methode van optimaliseren en toepassen van handelsregels aan dat technische analyse een out-of-sample voorspellende kracht heeft. Bovendien toont het schatten van Sharpe-Lintner CAPMs aan dat de beste recursieve methode van optimaliseren en toepassen van technische handelsstrategie¨ en een statistische signiÞcante voor risico gecorrigeerde voorspelkracht heeft voor ongeveer 40% van de onderzochte datareeksen, na correctie voor 0.10% transactiekosten. Echter, als de kosten toenemen tot 0.50% per order, dan heeft de recursieve procedure van optimaliseren en toepassen van handelsregels geen statistische signiÞcante voorspellende kracht meer voor bijna alle onderzochte datareeksen. Er kan dus worden geconcludeerd dat technische analyse winstgevend is en een statistische signiÞcante voorspellende kracht heeft voor een groep van aandelen genoteerd in de AEX-index, alleen als de transactiekosten voldoende laag zijn. In hoofdstuk 5 wordt de verzameling van 787 technische handelsstrategie¨ en uit hoofdstuk 3 toegepast op 50 indices van aandelenmarkten in Afrika, Azi¨ e, Europa, het Midden Oosten, Noord en Zuid Amerika en Oceani¨ e, en op de MSCI Wereld Index in de periode van januari 1981 tot en met juni 2002. Alhoewel de helft van de indices een continue investering tegen een rentevoet niet kon verslaan, wordt er net als in de hoofdstukken 3 en 4 voor beide selectiecriteria gevonden dat voor elke index een technische handelsregel kan worden geselecteerd die de passieve strategie van eenmaal kopen en vasthouden kan verslaan, ook als er gecorrigeerd wordt voor transactiekosten. Bovendien wordt er voor de helft van de indices gevonden dat de beste strategie een statistische signiÞcante voor risico gecorrigeerde voorspellende kracht heeft, zelfs na correctie voor 1% transactiekosten. Echter, als er tevens wordt gecorrigeerd voor de zoektocht naar de beste strategie, dan verwerpen zowel de RC als de SPA-test bij 0.25% transactiekosten voor de meeste aandelenindices niet de nul hypothese dat de beste strategie geselecteerd door het gemiddeld rendement criterium geen superieure voorspellende kracht heeft. Net als in hoofdstuk 4 worden er andere resultaten gevonden voor het Sharpe ratio criterium: voor een kwart van de indices, voornamelijk die in Azi¨ e, wordt de nul hypothese van geen superieure voorspellende kracht dan wel verworpen. Ook de recursieve methode van optimaliseren en toepassen van technische handelsregels toont aan dat technische analyse out-of-sample voorspelwinsten kan genereren, voornamelijk voor aandelenindices in Azi¨ e, Latijns Amerika, het Midden Oosten en Rusland, zelfs na implementatie van transactiekosten. Echter, voor aandelenindices in de VS, Japan en de meeste West Europese landen is de recursieve methode van optimaliseren en toepassen van technische handelsregels niet winstgevend als er een klein beetje transactiekosten worden opgevoerd. 

 

Bat de trend-volgende handelsstrategie¨ en geen statistische signiÞcante voor risico gecorrigeerde out-of-sample voorspellende kracht vertonen voor bijna alle indices. Alleen voor voldoende lage transactiekosten (≤ 0.25% per order) wordt er een economische en statistische signiÞcante voor risico gecorrigeerde out-of-sample voorspellende kracht gevonden voor trend-volgende technische handelsstrategie¨ en toegepast op de indices van de aande-lenmarkten in Azi¨ e, Latijns Amerika, het Midden Oosten en Rusland. Uit de resultaten van hoofdstuk 2 kan worden geconcludeerd dat een toevallig samenspel van economische factoren er voor kan zorgen dat technische analyse een schijnbaar voorspellende kracht kan vertonen. Uit de resultaten van hoofdstuk 3 blijkt dat met het toepassen van technische analyse op de Amerikaanse aandelenmarkt geen statistisch signiÞcant bovengemiddeld rendement kan worden behaald. Aandelen op de Nederlandse aandelenmarkt lijken wel enigszins voorspelbaar te zijn, zo blijkt uit hoofdstuk 4, maar transactiekosten doen de meeste positieve resultaten teniet. Na correctie voor transactiekosten, risico en de zoektocht naar de beste strategie, wordt in hoofdstuk 5 aangetoond dat technische analyse winstgevend is en een statistische signiÞcante voor risico gecorrigeerde out-of-sample voorspellende kracht heeft in de opkomende markten in Azi¨ e, het Midden Oosten, Rusland en Zuid Amerika. Echter, dit geldt alleen voor voldoende lage transactiekosten. Namelijk, voor transactiekosten groter dan of gelijk aan 0.50% per order worden er geen tot weinig signiÞcante resultaten gevonden. De conclusie van dit proefschrift is dan ook dat voor alle onderzochte Þnanci¨ ele datareeksen de zwakke vorm van de EMH niet zondermeer verworpen kan worden door het toepassen van trend-volgende technische handelsstrategie¨ en, nadat er is gecorrigeerd voor voldoende transactiekosten, risico, de zoektocht naar de beste strategie en out-of-sample voorspellen. In hoofdstuk 6 wordt een theoretisch Þnancieel markt model ontwikkeld met heterogeen adaptief lerende beleggers. De beleggers kunnen kiezen uit een fundamentele en een technische handelsregel. De fundamentele regel voorspelt dat de koers met een bepaalde snelheid terugkeert naar de fundamentele of intrinsieke waarde, terwijl de technische handelsregel is gebaseerd op voortschrijdende gemiddelden. Het model in hoofdstuk 6 is een uitbreiding van het Brock en Hommes (1998) heterogene agenten model, omdat het aan de verzameling van voorspelregels waaruit de agenten kunnen kiezen een realistische technische handelsregel gebaseerd op voortschrijdende gemiddelden toevoegt. Het model wijkt af door de aanname van relatieve risico aversie, zodat beleggers die dezelfde voorspelregel kiezen hetzelfde percentage van hun vermogen investeren in het risicovolle goed. Het lokale dynamische gedrag van het model rond het fundamentele evenwicht wordt bestudeerd door het vari¨ eren van de waarden van de modelparameters. Een mix van theoretische en numerieke methoden wordt gebruikt om de dynamica te analyseren.

 

In het bijzonder wordt aangetoond dat het fundamentele evenwicht instabiel kan worden als gevolg van een Hopf bifurcatie. De interactie tussen fundamentalisten en technische analisten kan er dus toe leiden dat de koers afwijkt van de fundamentele waarde en grote schommelingen vertoont. In deze heterogene wereld zijn fundamentalisten niet in staat om technische analisten uit de markt te drijven, maar fundamentalisten en technische analisten blijven voor altijd naast elkaar bestaan en hun relatieve invloed varieert door de tijd.

 

Achelis, S.B., 8 Alexander, S.S., 11, 12, 24, 45, 95, 150, 191 Allen, H., 5, 95, 278 Andrada-F´ elix, J., 21, 187 Arditti, F.D., 15 Arthur, W.B., 237 Bartlett, M.S., 77, 99, 149, 190 Bass, A.B., 6 Benington, G.A., 14, 25 Bessembinder, H., 20, 21, 187 Blume, M.E., 12, 17, 95, 266 Bollerslev, T., 58, 63 Box, G., 98, 99, 148, 150, 189, 191 Brock, W.A., 18, 19, 20, 21, 22, 23, 24, 25, 31, 34, 35, 43, 49, 52, 96, 100, 104, 187, 237, 238, 239, 244, 270, 300 Bu?et, W., 3 Chan, K., 20, 21, 187 Chan, L., 21 Chang, P.H.K., 7, 24 Cheung, K.C., 22, 187 Cheung, Y., 5, 95 Chinn, M.D., 5, 95 Cootner, P., 27 Coutts, J.A., 22, 187 Cowles, A., 9, 10, 11, 95 Curcio, R., 23 De Long, J.B., 237 303 Dempsey, M., 20, 187 Detry, P.J., 23 Diebold, F.X., 25, 42, 77, 96, 98, 99, 102, 148, 149, 189, 190 Dodd, D.L., 2 Dooley, M.P., 16, 19, 108 Edwards, R, 5 Edwards, R.D., 8 Efron, B., 18, 59, 96 Fama, E.F., 1, 12, 15, 16, 17, 34, 95, 152 Farmer, J.D., 6, 237 Fern´ andez-Rodr´ õguez, F., 21, 187 Frankel, J.A., 237 Freedman, D., 18, 59, 96 French, K.R., 34 Friedman, M., 237 Friend, I., 266 Froot, K.A., 237 Gaunersdorfer, A., 238 Goodhart, G., 23 Graham, B., 2 Grandia, V., 26 Gregoire, P., 23 Grossman, S.J., 266 Guillaume, D., 23 Hamilton, W.P., 4, 5, 10, 94 Hansen, P.R., 26, 29, 96, 104, 105, 110, 115, 118, 155, 161, 188, 198, 206, 208, 209, 297 Hollistein, M., 21.

 

304 Hommes, C.H., 31, 237, 238, 239, 244, 270, 300 Hong, H., 237 Hsieh, D.A., 99 Hudson, R., 20, 187 Isakov, D., 21 James, F.E., 14 Jensen, M.C., 14, 25 Keasey, K., 20, 187 Kendall, M.G., 11 Keynes, J.M., 3 Kho, B.C., 20 Kirman, A., 237 Knez, P., 21 Krishnan Guru, B., 23 Lakonishok, J., 18, 21, 34, 43, 96, 100, 187 Langedijk, N., 23 Leal, R.P.C., 21, 187 LeBaron, B., 18, 22, 34, 43, 96, 100, 187, 237 Lee, C.I., 19, 115 Leenders, C.T., 12, 152 Levich, R.M., 19, 34, 43, 52, 100 Levy, R.A., 14, 24, 25 Lintner, J., 13 Ljung, G., 98, 99, 148, 150, 189, 191 Lo, A.W., 7, 24, 34, 43 Lux, T., 237 MacKinlay, A.C., 34 Magee, J., 5, 8 Malkiel, B.G., 8, 9 Mamaysky, H., 7, 24, 43 Mandelbrot, B., 11 Mariano, R.S., 25, 96, 102 Markowitz, H.M., 13 Marquering, W., 115 Mat Nor, F., 23 Mathur, I., 19, 115 McCollough, W.A., 15 Menkho?, L., 5, 95 Mills, T.C., 20, 187 Ming Ming, L., 23 Murphy, J.J., 5, 38, 42 Neftci, S.N., 7, 24.

 

Nelson, D.B., 57 Newey, W., 108, 117, 153, 159, 195, 203 Osler, C.L., 7, 24 Payne, R., 23 Pesaran, M.H., 34, 115 Peters, S., 18, 59, 96 Pierce, D., 98, 99, 148, 150, 189, 191 Politis, D., 25, 103 Pring, M., 8, 42, 247 Ratner, M., 21, 187 Ready, M.J., 21, 23 Rhea, R., 4, 5, 95 Roberts, H.V., 11 Romano, J., 25, 103 Samuelson, P., 1, 34 Satterthwhaite, R.E., 49 Schabacker, R., 5 Schulmeister, S., 17 Shafer, J.R., 16, 19, 108 Sharpe, W., 13 Shiller, R.J., 266 Sosvilla-Rivero, S., 21, 187 Stein, J.C., 237 Sullivan, R., 25, 35, 43, 52, 89, 96, 100, 103, 104, 115, 187 Sweeney, R.J., 17 Taylor, M.P., 5, 95, 278.http://www.ttc-switch.com/Thesis_Writing/Finance/ 

 

The Tinbergen Institute is the Institute for Economic Research, which was founded in 1987 by the Faculties of Economics and Econometrics of the Erasmus Universiteit Rotterdam, Universiteit van Amsterdam and Vrije Universiteit Amsterdam. The Institute is named after the late Professor Jan Tinbergen, Dutch Nobel Prize laureate in economics in 1969. The Tinbergen Institute is located in Amsterdam and Rotterdam. The following books recently appeared in the Tinbergen Institute Research Series: 250. A.B. BERKELAAR, Strategic asset allocation and asset pricing. 251. B.J. VAN PRAAG, Earnings management; Empirical evidence on value relevance and income smoothing. 252. E. PEEK, Discretion in Þnancial reporting and properties of analysts’ earnings forecasts. 253. N. JONKER, Job performance and career prospects of auditors. 254. M.J.G. BUN, Accurate statistical analysis in dynamic panel data models. 255. P.C. VERHOEF, Analyzing customer relationships: Linking relational constructs and marketing instruments to customer behavior. 256. C.S. BOS, Time varying parameter models for inßation and exchange rates. 257. A. HEYMA, Dynamic models of labour force retirement; An empirical analysis of early exit in the Netherlands. 258. S. DEZELAN, The impact of institutional investors on equity markets and their liquidity. 259. D.J. DEKKER, Network perspectives on tasks in account management. 260. F.G. VAN OORT, Agglomeration, economic growth and innovation. Spatial analyses of knowledge externalities in the Netherlands. 261. U. KOCK, Social beneÞts and the ßow approach to the labor market. 262. D.J. BEZEMER, Structural change in Central European agriculture. Studies from the Czech and Slovak Republics. 263. D.P.J. BOTMAN, Globalization, heterogeneity, and imperfect information. 264. H.C. VAN DER BLONK, Changing the order, ordering the change. The evolution of MIS at the Dutch railways. 265. K. GERXHANI, The informal sector in transition. Tax evasion in an institutional vacuum. 266. R.A.J. BOSMAN, Emotions and economic behavior. An experimental investigation. 267. A.P. VAN VUUREN, Empirical analysis of job search using novel types of data. 268. H. VAN DE VELDEN, An experimental approach to expectation formation in dynamic economic systems. 269. L. MOERS, Institution, economic performance and transition. 270. N.K. BOOTS, Rare event simulation in models with heavy-tailed random variables. 271. P.J.M. MEERSMANS, Optimization of container handling systems. 272. J.G. VAN ROOIJEN, Flexibility in Þnancial accounting; income strategies and earnings management in the Netherlands. 273. D. ARNOLDUS, Family, family Þrm, and strategy. Six Dutch family Þrms in the food industry 1880-1970. 274. J.-P.P.E.F. BOSELIE, Human resource management, work systems and performance: A theoreticalempirical approach.

 

275. V.A. KARAMYCHEV, Essays on adverse selection: A dynamic perspective. 276. A.J. MENKVELD, Fragmented markets: Trading and price discovery. 277. D. ZEROM GODEFAY, Nonparametric prediction: Some selected topics. 278. T. DE GRAAFF, Migration, ethnic minorities and network externalities. 279. A. ZORLU, Absorption of immigrants in European labour markets. The Netherlands, United Kingdom and Norway. 280. B. JACOBS, Public Þnance and human capital. 281. PH. CUMPERAYOT, International Þnancial markets: Risk and extremes. 282. E.M. BAZSA-OLDENKAMP, Decision support for inventory systems with complete backlogging. 283. M.A.J. THEEBE, Housing market risks. 284. V. SADIRAJ, Essays on political and experimental economics. 285. J. LOEF, Incongruity between ads and consumer expectations of advertising. 286. J.J.J. JONKER, Target selection and optimal mail strategy in direct marketing. 287. S. CASERTA, Extreme values in auctions and risk analysis. 288. W.H. DAAL, A term structure model of interest rates and forward premia: An alternative approach. 289. H.K. CHAO, Representation and structure: The methodology of econometric models of consumption. 290. J. DALHUISEN, The economics of sustainable water use. Comparisons and lessons from urban areas. 291. P. DE BRUIN, Essays on modeling nonlinear time series. 292. J. ARDTS, All is well that begins well: A longitudinal study of organisational socialisation. 293. J.E.M. VAN NIEROP, Advanced choice models. 294. D.J. VAN VUUREN, The market for passenger transport by train. An empirical analysis. 295. A. FERRER CARBONELL, Quantitative analysis of well-being with economic applications. 296. L.M. VINHAS DE SOUZA, Beyond transition: Essays on the monetary integration of the accession countries in Eastern Europe. 297. J. LEVIN, Essays in the economics of education. 298. E. WIERSMA, Non-Þnancial performance measures: An empirical analysis of a change in a Þrm’s performance measurement system. 299. M. MEKONNEN AKALU, Projects for shareholder value: A capital budgeting perspective. 300. S. ROSSETTO, Optimal timing of strategic Þnancial decisions. 301. P.W. VAN FOREEST, Essays in Þnancial economics. 302. A. SIEGMANN, Optimal Þnancial decision making under loss averse preferences. 303. A. VAN DER HORST, Government interference in a dynamic economy. 304. A.P. RUSSO, The sustainable development of heritage cities and their regions: Analysis, Policy, Governance.