On Compounding
Designing a very profitable trading system is all about compounding. And if there is one thing that any trading method should strive for is to acquire, as much as possible, long term sustainable alpha points. Playing for a 40% return in one year has little value if it is lost the year after ($1.00 x 1.40 x (1 - 0.40) = $0.84).
The following graph shows $1,000 at various rates of return over 40 years:
The first observation should be that the first years do not seem to count for much; all the power is in the last few years of the series. But to get there, one has to survive all the previous years and maintain the higher rate of return, or, on average, its equivalent. There are only a few that operate, long term, above the 20% compounding rate of return. So you can be assured that each additional point is hard to get.
For comparison, the same graph as above but over 20 years results in:
Again the same curves, same shape, same exponential growth. The only real change is the scale, the outcome. And here again, it is the last few years that make all the difference.
The point I would like to make is: the stock market game is a long term endeavor.
There is nothing wrong with looking at the game or playing the game with a short term view of performance. But, IMHO, that is not where it really counts. You still have to look at the short term for the simple reason that that's where you have to trade, but it should also be looked at with a long term view of what your trading methods can do. And I think the only way to do this is test your trading methods over long term horizons. If your long term backtests do not produce what you are looking for, or at a minimum beat the Buy & Hold, then there should be no need to execute your strategy live going forward.
Before aiming for the 40 years time span, we should start with the first 20 and what to consider as the objectives of these trading strategies. How do you expand your initial trading capital to reach these objectives? And this all goes back to what kind of trading strategy will you employ?
I usually express it as a payoff matrix: Σ(H.*ΔP). In this one expression all is said as it does represent all the profits generated by your trading strategy over your time span. And since future price series will be the same for all, all that is left is to design better trading strategies H^{+}. There are thousands upon thousands of trading strategies out there; and a lot of these are positive. I classify them as H^{+}, my lazy way of saying exceeding the Buy & Hold strategy. By how much depends on the trading strategy itself.
It is what you put in your H^{+} that makes the difference. It translates your ability to recognize what is there and how to exploit it (meaning generate profits). The market does not know how you want to behave and technically doesn't care. Your trading counterparts have no feelings for your particular needs or status, their own objective is simply to get you (meaning your stake) since that is how they make their profits.
Designing an H^{+} strategy can be relatively easy. Over the past year I have used various starting points (from already published trading strategy) upon which I opted to improve their design by morphing their strategy design to my long term vision and involvement with the market. Some of my simulations can be found under the simulation menu. All these simulations exceeded my theoretical expectations as expressed in my first 2007 research paper: Alpha Power.
All that is said in my research papers is that one can improve performance by looking at the long term. And by doing so, one will start to consider that it is not the next trade that is important, it is the ensemble of all trades to be taken over the trading interval. And should you not want to stay the course, save yourself and do not start.
Created... August 19, 2012 © Guy R. Fleury. All rights reserved