System Trading Blog

Make Your Money Work Harder

Archive for December 2009

Week 51: Test account down 2.1% for the week. It´s more than just a headline figure.

leave a comment »

As empirical evidence suggests, some of the top-rated systems on zulutrade.com are money loosers in live accounts and zulutrade stats are misleading, to say the least.

Here is an example. A trading system Isunia.

This is what the performance graph shows on Zulutrade.

… and this is what it looks in real-life trading

Of course, there is a disclaimer — the account was not enabled for no limit trades. It was enabled for 3, 9 or 10 max open trades at any time. Still, during the month there are 90 trades executed into the account to make the findings statistically significant. There are no outliers, this is a steady, systematic decline of the account equity.

I removed Isunia from my test account system mix. Also, for the same reason, I removed monte financial and hii as two other smooth performance curve systems. They showed a marginal profit of 1 and 3 pips, but my conclusion was that, due to their style of trading, those systems will fail eventually.

Basically, those systems take risk of 350-400 pips or more per trade to go for the average gain of 15-20 pips, often averaging loss. There is one system I’ve left in the system mix for the test purposes, it’s named fx168fx. I was aware of the fact that the system is not free from the sins of averaging or taking indefinate loss for a limited gain, but it seemed to happen very rarely. However, now it seems that averaging down is systematic.

As an example, I give the current log of the test account.

He keeps buying GBPUSD in what is a strong downtrend.

…. but he keeps booking profits … ok, usually 1 pip at a time, but still profits … at least, before real-life slippage is accounted for …

and it’s consistently profitable …

… no wonder, the system is ranked #2 un zulutrade as at the time of writing …

So, should I be grateful and happily stay as a subscribe to the system. My answer is definate No. I remove it.

His stragegy looks like averaging the loss. It´s like doubling bet after every loosing roulette bet. Eventually, none has sufficient capital to execute the strategy for any substantial period of time. As I´ve heard in London City ¨averaging has killed more people than Hitler¨.

As the stats show, the guy takes more risk in one day (or even a single trade) than he books in profits for the whole month. My bet is that the strategy will fail, eventually.

P.S. As a hindsight, if I remember fen shui meaning of number 168, then it symbolises eternal wealth. For the trader, may be … not for the customer, definately. The guy is Chinese, it´s likely he knows the meaning.

Advertisements

Written by A.S.

December 18, 2009 at 9:03 pm

To risk 150 pips for a “sure gain” of 1?

leave a comment »

I test run a portfolio or several systems. Every system is included in the portfolio because of presumably good risk/reward ratio.

Today in the test account there were 5 “ugly” trades from 2 system providers. All trades profitable. The catch is that the profit was 1-6 pips per trade, but at some point the trades were in deep loss [130+ pips] before they were closed. Some of the long trades were done while market was in the middle of steep decline. Basically, idea is to buy whenever and hold until the position turns profitable.

Undoubtly, this is good way to produce smooth performance curve on zulutrade (zulutrade doesn’t record intraday moves, but logs performance data on trade-close basis), but it is very risky trading strategy.

Written by A.S.

December 7, 2009 at 11:30 pm

Posted in Uncategorized

Week 42 Test Account Review: up 0.3%

leave a comment »

The account is up [just] 0.3%. One system is removed from the test as it apparently lacks any sensible risk control measures.

Written by A.S.

December 4, 2009 at 9:27 pm

Posted in Uncategorized