"The problem is all inside your head/ she said to me" is how Paul Simon starts his 1975 hit, for there must be fifty ways to leave your lover, he speculates. He was not crooning about trading, but he might well have been. Unless you are a systematic and highly disciplined trader, you have probably discovered fifty ways to exit a trade. Well if not fifty, surely more than ten. The actual effect is to shorten the length of your average trade and thus possibly degrade your ratio of average winning trade to average losing trade. But why is this significant? Let me offer a simplified explanation.
TOPS COLA or the Case for Trend Following
"Buy Low, Sell High", the old Wall Street maxim is a prescription for trend-following without a trailing stop. I prefer the strategic formulation "TOPS COLA", for take our profits slowly, but cut off losses at once.
It is easy to show that the average absolute close-to-close percentage change in a random stock on average increases at the holding period increases (see Chart 1). The universe of stocks in Chart 1 is those in the S&P 500, from roughly 2004 to 2017, with sample size ranging from about 535,000 at the 3-day holding period to approximately 416,000 at the 250-day period. Here the length of the holding period is in trading days, not calendar days. So the idea that we should extend our holding period to improve the average profit is well supported.
Chart 1: With sample size ranging from about 535,000 at the 3-day period to about 416,000 at the 250-day period, data from the S&P 500 universe of stocks show that extending the holding period also extends the range of the movement experienced in the typical stock. In other words, trend-following should typically yield greater average return as we extend the holding period. The standard deviation of the expected range decreases from left to right on the chart, i.e., noise decreases. (Note however, this chart is not a system, i.e., it does not tell you what your average losing trade might be, nor the percentage of winning trades.)
Since trend-following with a trailing stop makes no assumptions about the amplitude of the move (no profit targets) or the duration of the move (no time stops), we are likely to participate in a long-duration trend without succumbing to an arbitrary exit. The trailing stop has the effect of cutting off losing trades relatively early, but letting profitable trades run (TOPS COLA). Since only a handful of trades are truly awesome trends, in general, the trend-following strategy relies on outlier trades to boost the ratio of average win/loss trade (see below).
Fortunately, there is no upper limit to the price gain should your holding period extend to "infinity". Unfortunately, trend-following is hard to implement day-in and day-out because we know neither the trade that will have the large move, nor the exact sequence in which trades will arrive. It is precisely this uncertainty which makes following a system contrary to human nature (which, paradoxically, is the strongest reason to do it).
Exits for Systematic Traders
The problem for humans is dealing with our emotions. Here are three typical exit strategies for systematic traders that have emotional repercussions.
A trailing stop: You follow the trend and exit when the market goes through your trailing stop. You will probably have regrets, since the stops always seem to move at the wrong speed, either too fast or too slow.
A profit target (limit): You exit when your profit target is hit. You will be remorseful when the stock continues to move along its prior path.
A time stop: You exit on the close or open after a set time in trade. You will hate it when the market suddenly accelerates in the direction of the trade you just exited.
Cognitive Biases Push Us toward Early Exits
We can be easily affected by cognitive biases. For example, loss aversion, when emotional pain of losing $1 is worse than the joy of making $1, is one powerful tendency pushing us to cut trades short. A squall of selling, accompanied by negative comments from analysts or other perceived experts could also force premature exits, a type of confirmation bias, if you thought the market or stock was overvalued or overbought. The other line in the supermarket always moves faster, or so it seems. Thus, boredom becomes a problem, as stocks in your portfolio languish, but those not in your portfolios flourish, causing you to exit. The once-burned-twice-shy effect (the availability heuristic), may cause you to avoid or hasten trades in certain sectors. You might hold on to a large losing trade while selling out a small wining trade, even though the two trades are independent. I could go on, but you get the idea.
Why does it Matter?
The math of net trading profits is quite simple:
Net Profit($) = # of Winning Trades*Average Winning Trade($) - # of Losing Trades*Average Losing Trade($)
This equation says that we will be profitable over the long run by maximizing the proportion of winning trades and the ratio of average winning trade (AWT) to average losing trade (ALT). In fact, a robust, well designed trading system will do just that over the test data. Hence, by following a system, we increase the odds of approaching the performance in the historical test. However, back testing apart, the real time performance of a system could be below expectations over any given period, and so one has to keep trading a system through ups and downs, which can be difficult. If we are arbitrary, then the realized winning percentage as well as the ratio of average winning to losing trade could become unpredictable, and we will deviate from our testing. So the hard part is to keep the emotions out of it over the long run, which explains in part the desirability of short-term trading.
How to Evaluate Your Trading
The two ratios you need to know accurately are as follows:
1) The percentage of winning trades:
% Win = 100*(# of Winning Trades/(# of Winning Trades + # of Losing Trades))
2) Average Win/Loss Ratio (AWL)
AWT = Average Winning Trade = (Gross Profit/# of Winning Trades)
ALT = Average Losing Trade = (Gross Loss/# of Losing Trades)
AWL = Average Win/Loss ratio = AWT/ALT
Your goal it to maximize both ratios. So, for example, you do not want many small winners and a few large losing trades, so that %Win is high, but AWL is low. Conversely, a few trades could be inflating the AWL, when the percentage of winning trades is small (say less than 25%).
Subtle Change in the Markets
There is a subtle change in the markets. As stocks have moved to new highs, defensive sectors, bonds and gold have flattened or turned lower (see Chart 2). The intermediate-term and long-term stock trend remains higher, as it has throughout this year.
Chart 2: As highlighted in the red box, bonds have turned lower in the short-term, whereas gold has turned neutral. The stock market is trending higher on all time frames, and the dollar is trending lower. For a review of my trend-following models, follow this link.
The first is to calculate the two ratios above. The second is to listen to "50 Ways to Leave Your Lover" by Paul Simon and give me some suggestions for an alternative chorus, so that it relates to trading. I just loved the chorus and its rhyming scheme, which was apparently rooted in nursery rhymes for 5-year-olds. With profuse apologies to Paul Simon, here is my feeble suggestion.
Don't fade the trade, Jade
Just move your stop, Pop
Ignore the stress, Jess
And let your trade run
When the market's trendy, Wendy
Ignore the vol, Sol
Just be wary, Gary
But set your trade free
Thank you for tuning in, and I hope you will subscribe to the blog using the quick link below.