Combining the results of DLPAL LS and DLPAL DQ to improve the performance of equity long/short and directional strategies based on conflicting and confirming signals.
DLPAL LS is unique software that calculates features measuring the directional bias of securities and also generates historical values of those features. The software is used by professional traders and hedge funds for the development of directional and long/short trading strategies based on fixed algos or machine learning using generated train and score files.
DLPAL DQ is an end-of-day scanner for discretionary quantitative traders. The software scans any number of securities for price action anomalies. Validation tools are included and code is generated for a variety of platforms for further testing.
Note that DLPAL DQ uses a much richer set of features (Major Feature Cluster) while DLPAL LS uses a minimal set. Therefore, DLPAL DQ results can provide additional information that can be used alongside results obtained by DLPAL LS. Specifically, signals from DLPAL DQ may contradict or confirm signals from DLPAL LS. This information may be used for risk management purposes. Below are specific examples from running DLPAL LS and DLPAL DQ with a universe of 84 large caps as of the close of March 27, 2019. In all images below DLPAL LS results are on the left and DLPAL DQ on the right. The latter results include the return from open to close of the day for each stock under Return column.
CAT signal
DLPAL LS indicated negative bias for CAT while DLPAL DQ identified a long strategy. CAT gained 0.356% from open to close of the following day. Therefore, ignoring the negative bias in DLPAL LS results for CAT based on DLPAL DQ validation would have paid off.
GD signal
DLPAL LS indicated negative bias for GD while DLPAL DQ identified a long strategy. GD gained 0.341% from open to close of the following day. Therefore, ignoring the negative bias in DLPAL LS results for GD based on DLPAL DQ validation would have paid off.
MCD signal
DLPAL LS indicated marginally negative bias for MCD while DLPAL DQ identified a long strategy. MCD gained 0.755% from open to close of the following day. Therefore, ignoring the negative bias in DLPAL LS results for MCD based on DLPAL DQ validation would have paid off.
XOM multiple signals
DLPAL LS indicated positive bias for XOM while DLPAL DQ generated multiple long signals for the same security. Normally, multiple signals increase the probability of success. This information can be used for risk and money management purposes. XOM gained 0.887% from the open to the close of the next day.
Many possibilities
Our customers continuously amaze us by discovering new ways of using our software tools. From a user several years ago who developed a Bayesian analysis framework based on DLPAL S strategy signals to a user who used the signals with a genetic programming engine to contract the optimum strategy now to a user who has developed a way of combining the signals of DLPAL LS and DLPAL DQ in an attempt to arbitrage conflicts or multiple signals.
For articles with examples using all three versions of DLPAL click here.
Demo trials
Click here to order demo trials.
