My pleasure! :)

Excellent article! Thank you!

Interesting. I'm used to thinking in terms of track record rather than analysing trades. That's because I trade pretty slowly, and with a trend following system the # of wins would be less than 50% even with a profitable system.

Certainly the track record you showed in isolation wouldn't pass a test for significance based on say daily returns.

One point is that you don' t analyse the size of the losing trade vs the winning trade. I guess the better test is to look at the average profit per trade (which encapsulates both win rate and win:loss ratio) and determine if that is different from zero.

Still with only 3 years and a high win:loss ratio (indicating it's probably some kind of mean reversion or option selling strategy) I'd be concerned about a 'peso problem'.

Robert Carver,

yes, you are right that my approach is not (equally good) applicable to trade following strategies (with relatively few trades). However, let me see the trade history and I will find a way. For trendfollower I would analyse what part of trend s/he managed to exploit, how his/her strategies performs when the main indices do not trend and so on.

>One point is that you don' t analyse the size of the losing trade vs the winning trade.

I do! (Figures 2 und 2a)

Moreover, I emphasize that he trades virtually without a drawdown.

>option selling strategy

No-no :). You see, I know all assets (ISINs) which he trade. This statistics is directly from wikifolio, not from him, so it is trustworthy.

Moreover, he, himself, recommends (to retail investors) never trade with options, CFDs and other leveraged financial products.

Vasily, one has to agree. Those are really impressive numbers. It represents a CAGR in excess of 250%.

From what I have seen, it is a discretionary system. Do you know anything about attempts to automate that trading strategy or something similar?

The math is incorrect. This would be a hypothesis test with a slope that subtracts the current market appreciation / depreciations since 2013. It's like trying to predict if it will be sunny in California.

The results seem incredibly impressive (and I don't want to take away from what appears to be a great job), but the statistics used to analyze the results are wrong.

Most successful prop traders accomplish this task and good luck scaling this type of return with decent real capital, only a very few are capable, but that doesn't mean returns like this can't be done, scale and talent are obstacles.

Guy R. Fleury

>Do you know anything about attempts to automate that trading strategy or something similar?

Unfortunately no. Moreover, I explicitely know that, as you say, he trades essentially discretional (he comments his trades very actively, however, in German :)). In particular, he looks for acquisition candidates, quarter reports and so on. Though he seems to use some technical indicators as well, they are secondary for him.

Steve Dwek, CIM,

>The math is incorrect...

Can you please explain it more in detail? To be true, I don't understand what do you mean with "a slope that subtracts the current market appreciation"

Mike Agne,

Yes, scaling is a big problem. Einstein, himself, admits that his strategy is non-scalable for €1 Mio (let alone €1 Bio).

So for a mutual fund there would be insufficient market liquidity, however, for a typical private investor the strategy is still applicable.

Vasily, thanks for the clarification. It requires a very good understanding of markets and how they operate to achieve those numbers on a discretionary basis. If you ever see him, send him my congratulations for more than a great job. Those are fabulous numbers.

After having looked at the trades, I can understand why the methodology is not that scalable. But, as you said, a trader that starts small, needs in the beginning, to go faster than most.

of course one can beat the market provided the 95 pct of competition lacks patience, stamina and courage. like in any other professional sport