Backtesting

I was a fan of backtesting when I started looking at trading as it gave me some cool numbers to indicate what I "could" have profited had I followed a particular strategy. I learned a few things about the fallacy of backtesting in the process, namely that it is historical and can be misleading.

The key is that the testing proves that a strategy has potential and needs to be followed up with live testing.

I did a bit of historical testing of what I am calling PP200, just 8 trading days worth.

I used the PP, R and S lines primarily, the 200sma for some reference and the 1/2 lines occasionally. I placed the orders on the first pullback to test the respective lines and exited on a re-crossing, weakening of the price, next line hit or, if the move had a large momentum then when the price move weakened after the next target was crossed. Orders were only placed if the pullback came to within 2 cents of the target line to give me a mechanical entry criteria. Price used was the target plus 2 cents for the entry and minus 2 cents on the exit. Trading was stopped every day at 1300h. Commissions are accounted for in the net profits.


There were very few losing trades, which surprised me as I expected a fair number, perhaps as high as 40%. I only saw 6%, so I will assume that is higher in real life. There were fifty trades altogether.

TLM, long only trades

Average daily return (based on the full $4K portfolio) was 1.45%
Total for the period was 11.6%

HED, long only trades

Average daily return (based on the full $4K portfolio) was 2.68%
Total for the period was 21.45%

$1322 total combined net return or 33%.

My normal fudge is to drop the net return by another 30%, so that makes $925.40 or 23.1%

Interestingly the HED trades only number 16 but almost doubled the returns from 34 trades of TLM. I think this may be due to the fact that HED follows the capped energy index and smooths out the variances and hiccups along the way whereas TLM is a straight up stock. I believe that this will require some testing under fire as this mitigation may prove useful for smoother trading results.

Either way that is a decent return for two weeks trading. One of these days I will get around to applying this strategy to prove the results.

I keep in mind that all of the live testing that I have done in the past was with real money and I had no qualms about losing a bit here and there in the interest of learning the ropes. This month I have been very hesitant to pull the trigger even though I know my strategy. It would appear that "getting serious" has some psychological implications that I did not anticipate.

Jeff.