Musings on Technical Analysis

I frequent a trading related chat-room, the primary purpose of which is to relate market news and fundamentals however there is one trader in the room that had been sharing his trades in real-time and primarily utilized MACD crossover’s for trade entry. The other criteria, though not stated initially, was focus only on highly volatile stocks; something either exploding higher or imploding lower. When I started systematic back-testing, I tested most of the simple, freely available technical studies and wrote off 99% of them when I realized they didn’t work over a large sample. So I was quite hesitant at first to give this trader or his system any additional thought. But his track record, which he posted each day was undeniable. He was consistently making money each day and that fact alone forced me to take a closer look.

So I begun some experiments myself with MACD. I began with a systematic back-test and as figured, the results were awful. However, I forged ahead anyway as I noticed visually, quite a few profitable trades with attractive risk rewards. The problem, as with any trend-following system, was the number of whipsaws. So I decided to forward test instead, using 100 shares, something small enough where the P&L would have no impact and it was quite easy to manually execute. As usual I keep track of all my statistics. After a sample of 48 trades, I had a 63% win rate and ~.75 avg win to avg loss (unfortunately my avg loss was bigger then my avg gain) yielding a net positive P&L. I should point out that as the test continued, I decided to increase the size and add more discretion to the exits, which increased my win rate however decreased the payout. This is an aspect of the system that I need to work on. Additionally I should point out my commission rates are likely higher then most and this has quite an adverse effect on payout ratios.

Here’s an example from Friday, 3 nice trades in VXX, a trading vehicle I particularly like due to it’s recent volatility and my current trading thesis on the name itself:

VXX is a nice trading vehicle for this strategy

Now MACD is nothing special, in fact it’s simply a derivative of a moving average cross-over system. In fact, an MA system would probably be just as profitable so there is no need really for MACD in particular. The profitability of the strategy is contingent upon finding the right security to trade. Let me say that again, if someone is considering utilizing this strategy, your profitability will be determined by the stocks you pick rather then following the model blindly.

So why MACD if it’s nothing special? Well for one, someone else is using it profitably so psychologically I have a profitable model to replicate. More importantly for a discretionary strategy, it immediately provides a visual structure to my trades, allowing me to know my entry and stop loss point and quickly size the trade based on risk and reward. Having that structure prevents me from say picking a top or catching a falling knife, it provides a stark visual cue to remind me when I’m wrong and exit a trade. It gives me a cue to re-enter a trade after a losing trade. In summation, it helps me avoid mistakes more then anything. Execution is of paramount importance in prop trading, good ideas are plenty but it is meaningless without proper execution.

And let me also say that it is quite easy to reconcile the different conclusions rendered between the poor back-test and the profitable forward test. In any scenario, I would place far more importance on a real-time, forward test then back-test. Especially since it’s with real money and executions. Back-test provide a good place to start but the key in strategy development is the forward test. Even if one doesn’t have the ability to program a proper back-test, one can take any visual gleamed rule and test it on a small basis or even paper trade to find a result.

Nice trades in AAPL on recent reversal day




~ by largecaptrader on February 19, 2012.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: