Why Do Laurence Connors Market Breadth Trading Strategies Stopped Working?

The interest of using market breadth to time the markets can be attributed to one person, Laurence Connors, thanks to his influence book Investment Secrets of a Hedge Fund Manager. The book was published back in year 1995. The strategies explained in the book are revolutionary because they are all complete mechanical trading strategies that are backtested against many years of historical data. Since then many generations of traders are inspired by Mr. Connors’ ideas to research on market breadth and build their own trading strategies around them. However, majority of these efforts are not resulting in anything functional.

I am going to explain why these efforts have all failed and how not to design trading strategies that are destine to fail in the future.

Case Study: Connors Hayward Advance Decline Trading Pattern

I have written about this strategy featured in the book many years ago in CHADTP Past and Present. The article was written back in year 2013 and I have not posted any update since. Here is the chart posted back then with the equity curve at the bottom.

Following is the equity curve of the same strategy implemented in Tradestation, updated to include everything up to Jun 2020. The number of days to exit is set to 7 which matches the green equity curve on the chart above.

Notice the equity curve matches perfectly so we know the strategy implementation is correct.

The troublesome issue is that the strategy failed to produce any more profit since year 2013 and basically falling apart since year 2016.

Well, the strategy has a great run since year 1995 up to year 2013. It proves the concept has validity. However, it also sends a chilling message to all those who research on trading strategy that, maybe, all strategies will fall apart some day.

My Take of the Core Problems with CHADTP

Mr. Connors insight into using advance decline issues is a great one. The discovery of the mechanical strategy is a great accomplishment. However, there are several underlying issues with the way he was using the information that caused the strategy to eventually breaks down.

First, I wrote many years ago that Net Advance Decline Issues is a bad idea in A Robust Method To Interpret Advance Decline Issues. Due to the instability of the data, any trading strategy designed with this as the core driver will suffer in the future. The proper way to utilize advance decline data is by normalizing the information into advance issues percentage. All the details is explained in that article so I won’t repeat it here.

Second, is the fatal mistake of picking top and bottom without knowing its dependency of optimized parameters. Majority of the trading strategies you see out there all suffer from this issue. That’s why they fall apart when you try to use them in your own trading.

Third, microstructure of intraday price behaviour is not properly utilized in the strategy.

Fourth, days to exit is a bias variable for all index futures.

Dependency on Hidden Optimized Factor

Using CHADTP as an example. Its core decision variable is the 5 days average of the Net Advance Decline Issues. The rule for going long is that when the variable value drops below –400, the strategy tries to go long using upside breakout as the confirmation of reversal.

In English, what the strategy is trying to do is that if the net advance decline issues has been pretty negative over a 5 day period, the strategy will find an excuse to go long, hopefully, you’ve picked the bottom. The short side approach is the exact opposite.

5 days of average value of Net Advance Decline Issues at –400 may be serious back before year 2000. That’s because the number of traded issues everyday back then was about 2500. So 5 x –400 = –2000 was a difficult value to drop down to. In other words, the 5 day average and –400 was the optimized result based on the distribution of Net Advance Decline Issues.

Nowadays, just a single day selloff can result in –2000 in Net Advance Decline Issues. Even if the 4 prior trading days has a net value of zero, the bottom picking will still be activated.

In principle, any form of top and bottom picking based on certain averages and specific threshold will always fall apart in the future as we cannot predict how the distribution will look like in the future.

The fundamental premise assumed to be true by this strategy has broken down for years. So, the breakdown of the strategy is the expected outcome.

I wrote all about this in Advance / Decline Issues Past and Present years ago.

Disrespect of Microstructure of Intraday Price Behaviour

The entry method used by the strategy is to stop itself into a position when the next day price print exceeds the signal day’s price extreme. The authors’ idea is that if the market manages to print outside of the signal day’s range, it is a confirmation of the market’s intention to reverse.

As I explained in STOPD, that these price extremes are always challenged and failure to clear them will lead to a move towards the other end. So the entry method really did nothing to help the strategy.

To proven my point, here is the equity curve of the same strategy that enter the market on close.

May be Mr. Connors want to hide this fact in plain sight so that he has less competition when he needs to get his orders filled? LOL

But even with this much improved result, the huge losses are still there. The weaknesses of the strategy cannot be overcame by just the entry method.

Proper Use of Days To Exit

Index futures are bullish bias by design. Hence extending the number of days to exit will automatically reward the long trades with more profit while hammering the short trades. In general it is important to separate the exit criteria for long positions and short positions because price does not move the same way when it is going up, comparing to the way it is moving down.

I understand the strategy is just for illustration purpose in the book. So readers of this article now know the importance of not sharing the exit conditions among long and short positions.

Anything We Can Do to Savage the Idea?

Yes and no.

Yes, that the strategy does not employ any price pattern as a filter which can make a world of difference. Using a reasonable money management scheme can help. And switching to advance issues percentage can definite make it all more stable.

No in the sense that if we do all of the above, it is no longer the original idea.

Afterthoughts

I have been doing research on market breadth analysis more than two decades now. It is not that difficult to create robust swing trading strategies using market breadth data. The main problem is that people cannot overcome their own weaknesses of wanting to see yet another top and bottom picking model to satisfy their insecurity. This leads to tunnel vision and inability to look at the problem the right way.

I explained in details on how to really fix yourself first before you approach trading in Trading Success Blueprint. Even the example trading strategies presented in the course are market breadth based. It is hard to engage the market until you have developed a healthy attitude towards trading. That is really what separate consistently profitable traders and struggling ones who keep looking for the best way to engage the market. In turn, it allows this strategy and its variations to stay alive and resurface every year among discussions by traders.

Part of Market Breadth Primer series

Share