# #077: PRINCIPLE OF "STATISTICAL ARBITRAGE" APPLIED ON MARKET INTERNALS

In one of my previous articles I focused on the concept of Market Quality created with the help of Market Internals. In today’s article I will get back to the original topic of Market Internals. I would like to present another interesting concept of their application.

One of the generally popular styles of trading is called Statistical Arbitrage (or StatArb). The principle of this trading style is very simple:

1. We take two highly correlated shares; often from the same industry – e.g. Pepsi (PEP) and Coca-Cola (KO).

2. These kinds of shares correlate very closely. Thanks to that, they are occasionally converging and sometimes diverging. These are great opportunities for trading. How?

3. First of all, we calculate the difference between either the prices of both shares, or their ratio.

4. Afterwards, we calculate the “usual” difference or ratio – the so-called MEAN. To do this, we use a simple moving average of the price difference or ratio.

5. Then we look for extremes, i.e. +2 standard deviations (STADEV) and -2 standard deviations from MEAN.

6. If the difference is above +2 STADEV, we sell share A and buy share B; if the difference is under -2 STADEV, then we sell share B and buy share A.

7. We simply speculate on the return of the difference/ratio of price back to its usual MEAN where we exit the trade.

For a better understanding, please have a look at the below images (previously used in a presentation I had abroad):

Shares KO – PEP. From the first point of view, it is obvious that both shares correlate very closely; they are occasionally converging and sometimes diverging:

The basics of StatArb trading is very simple – we create MEAN from DIFFERENCE or RATIO, we define +2/-2 STADEV, and, once exceeded, we buy and sell simultaneously. This is the reason why StatArb is also called “pairs trading”.

How can we apply this concept on Market Internals?

It’s actually rather simple:

1. Some Market Internals are also “paired”, e.g. UVOL-DVOL or ADVN-DECN.

2. It means we can create DIFFERENCE or RATIO from these pairs and, therefore, we can also create MEAN and +2/-2 STADEV.

3. The only difference is that instead of Market Internals which we can’t trade, we buy or sell the underlying asset.

Here is an example with a pair of UVOL-DVOL:

1. We create the difference for UVOL-DVOL.

2. We calculate MEAN (MovingAverage (DIFFERENCE) ) – the period is entirely up to you.

4. Once +2 STADEV is exceeded; the volume on the market as a whole is in favour of the LONG side. We can use this information to trade any stock index markets (ES, TF, YM, EMD, etc.).

5. Once -2 STADEV is exceeded; the volume on the market as a whole is in favour of the SHORT side. Again, we can use this information to trade any stock index markets.

And now: how can we make use of such information?

Again, there are many possibilities:

1. As a “Master-filter” to enter the trade (on stock index markets).

2. To increase/decrease contracts.

3. To tighten SL.

4. For early exits with part of our positions.

5. Anything creative you can come up with – it is necessary to think and experiment. :-)

The exceeding of standard deviation won’t always mean trend trading. For example, exceeding of +2 STADEV can mean (in certain situations) trend exhaustion and a possibility of correction – i.e. the opportunity for short-term trade against the trend.

There is a need to test each and every system, as there isn’t just one universal use. This concept is possible to use in many other ways.

There are many advanced ways to use this concept but that is a topic for some other time. Today was just a brief introduction so that the more aggressive and advanced traders can continue to work with this idea further.

Tomas

##### Author: Tom Nesnidal (more about me)
###### What others are saying

"Tomas is one of the most creative traders I know. He is able to generate unique trading ideas and elegant solutions to system development challenges."

"Tomas has been a professional trader for more than a decade and I have had the privilege of monitoring his accounts in action since 2006. His systems are performing as some of the best I have ever monitored and executed. I have been in the trading industry since 2002 and worked with many developers from all over the world. Tomas is definitely the one to consider."

Martin Lembak,

MFRM, CAIA

"Tomas is a professional trader, who for the last 10 years has specialized in developing trading systems. We have been tracking his trading systems for about 5 years and they generally show very robust, stable and above average performance. Striker is pleased to work with someone like him - a real professional with enthusiasm and deep knowledge of trading."

William Galwas, President Of Striker Securities, Inc

"Personal consultation with Tom helped me to re-evaluate the complexity of my robustness testing and optimization processes. Plus, it has helped me with a specific plan on how to take things forward. It has given me some great ideas on how to avoid overfitting and make my testing more robust and provided tips on low hanging fruit in terms of the best markets to trade for intraday/short-term breakouts."

Craig Peters,

United Kingdom

"Tom's approach to Automated Trading Strategies design, tests of robustness and portfolio diversification is really unique. He has been a professional trader for many years and the depth of his understanding of Breakout strategies and Market Internals is hard to find elsewhere."

Antonin Fisher,

Hedge Fund Manager,

Czech Republic

"Tom´s systems and trading approach do really work and can bring good, stable and reasonable returns. I can highly recommend him as a teacher."

David Hruby,

"Tomas is truly an inspiring mentor and great trader with a positive life attitude. He has got very deep understanding of markets, breakout strategies, robustness testing procedures, and Market Internals."

Jan Lesanovsky,

Client, United Kingdom

"As an experienced trader, I used Tomas´s consultation to get more clarity on best-practise robustness testing procedures, risk management, and portfolio management. He was quickly able to help me refine my current trading and system development approaches, and pointed me to really practical measures that I could implement immediately. I found Tomas to be not only extremely knowledgeable but very helpful and a great teacher. This has given me more confidence in developing robust strategies that work, and I would recommend his mentoring as extremely valuable for any systems trader."

Trevor West,