Advanced Trading Strategies
- Pitfalls in Optimizing Statistical Trading Strategies (Part II)
- Nasdaq forecaster
- Black box trading
- Resources: product reviews, seminars, web sites
- E-groups, webrings and Yahoo clubs
- Preview: in the next issues
- Nasdaq forecaster algorithm
- Subscribe / unsubscribe
Our team has been working hard to refine DataShaping statistical trading
strategies. We will soon offer enhanced keys, for the same price. These
keys rely on the latest technology described in our article below.
We are also working on automating order execution. Eventually, in
collaboration with partner brokers, we will offer a product that performs
fully automated trades every day. We are also in negotiation
with top financial portals to provide Nasdaq forecasts online.
Meanwhile, our website is being improved every week. New products and
services are added: check out our announcements section.
Pitfalls in Optimizing Statistical Trading Strategies.
Part II: Improving long-term return on short-term strategies.
In this article, we describe how to backtest a short-term strategy to
assess its long-term return distribution. We will focus on strategies
that require frequent updates. They are also called adaptive strategies.
We examine an undesirable long-term characteristic shared by many of
these systems: long-term oscillations with zero return on average. We
propose a solution that takes advantage of the periodic nature of the
return function, to design a truly profitable system.
When a strategy relies on parameters requiring frequent updates, one has
to design appropriate backtesting tools. From our
we know that we
should limit the number of parameters to six. We have also learned how to
improve backtesting techniques, using robust statistical methods and
constrained optimization. For the sake of simplicity, we assume that the
system to be tested provides daily signals and needs monthly updates. The
correct way to test such an adaptive system is to backtest it one month
at a time, on historical data, as follows.
For each month in the test period, do:
The whole test period should be at least 18 months long. Thus we need to
gather and process 24 months worth of historical data (18 months, plus 6
extra months for backtesting). Monthly returns obtained
sequentially OUT OF SAMPLE (one month at a time) in step 2 should be recorded for
further investigation. You are likely to observe the following patterns:
- Step 1: Backtesting
Collect the last six months worth of historical data prior the
month of interest. Backtest the system on these six months to
estimate the parameters of the model.
- Step 2: Walk forward
Apply the trading system with the parameters obtained in step 1 to
the month of interest. Compute the daily gains.
We have now all the ingredients to build a long term reliable system,
that we will call a metastrategy, since it is built on top of the
original system. It works as follows:
- many months are performing very well
- many months are performing very badly
- on average the return is zero
- good months are often followed by good months
- bad months are often followed by bad months
If last month return (as obtained in step 2) is positive, use strategy
this month, otherwise use reverse strategy by swapping buy and sell
This feature will be introduced in some of our selected trading keys,
resulting in a lower but more consistent long-term return. A special
symbol will be used to identify these high-tech keys.
A n n o u n c e m e n t s
- Nasdaq forecaster
It computes the
chance that Nasdaq (or any other index or stock) will move up tomorrow,
between open and close, by a given percentage. It is based on a non
parametric model and constitutes our first type of universal key.
The optimizer is a sophisticated program used to determine strategy
parameters that meet specific criteria. Professional and institutional
investors can contact us to obtain a customized version of the
optimizer. Our phone number is 925-432-4104. Please ask for
Vincent Granville, our Strategy Architect.
- Automated Trades
Our one-buy-one-sell daily system will be fully automated when "exact
or at close" limit orders are available. We are currently investigating
the feasibility with online and traditional brokers. If you offer broker
services and would like to partner with us and provide this type of limit
order, please contact us, at 925-432-4104. Ask for Vincent Granville, our
There are many thriving online communities focusing on stock trading,
some of them being well maintained and spam-free. Besides popular message
boards such as SiliconInvestor.com, there are Webrings (interesting if
you own a financial web site), Egroups and Yahoo clubs. The easiest way
to discover these resources is to use a search engine such as Google, and
type in ad-hoc keywords in the search box. Start with . The first search result is Ed Borasky's egroup,
one that I recommend. You can also check which clubs and egroups are most
visited. Useful links:
Note that my previous search asks Google to look for pages that contain
both "egroup" in one word and "computational finance" in one word as well.
It is a more efficient search than .
In the next issue of this newsletter, we will feature financial websites
that compliment Data Shaping.
A d v e r t i s e m e n t s
Daily stock picks for short term traders using Japanese Candlestick
Patterns. Free Play of the Day picks; Glossary of candlestick terms and
definitions; Traders Forum; Extensive list of stock picks based on daily
patterns; 30 day free trial.
Reduce your commissions substantially!!! Do not compromise, you can have
it all! Let Online Trading Academy's relationships with top direct access
brokerages help you get: Better commission prices, Better service on your
account, Direct access platforms, and Bulk discount on commissions!
To appear in the next issues:
To appear in the professional issue:
- Nasdaq forecaster: algorithm
- Trading strategies: pitfalls (III)
- Trading keys: long term return
- Best metrics to assess performance
- Updated list of stocks likely to outperform with our system
- Brokers offering automated trading systems
To subscribe or unsubscribe, go to our web site and fill out the
You can also use this form
to refer friends, business partners, colleagues or clients. This
newsletter is sent occasionally, when we release new products or when our
Research Department sends us an interesting article for publication.
Vincent Granville, Ph.D., Editor
Data Shaping Solutions