Can a consistently losing strategy be profitable if I do the opposite of it?
- Thread starter Amatrue
- Start date
IMHO: For any strategy: it is wise to understand why it works when it is successful as well as understand why it does not work when it fails.
Probably not in most cases. It is nearly as hard to lose on every trade as it is to win on every trade. And if the losses are because of b/a spread losses, commission, fat fingers, under capitalization, too much fear and greed, etc., then the same bad results will happen.
%%
Most likely would be profitable if you are on right side of the main trend +slippage is reasonable.
A consistanty losing plan on 1 or 5 minute charts will most likely not work, even if flipped;
many data feeds dont even work right on 5 minutes/get rich quick.
Trading for dividends as a #1 goal could work well if you did the opposite + sometimes collected dividends...…………………………………………………………………………………………………………...
BUY + hold inverse ETFs is a consistant loser;
long etfs works much better for longer periods/so it depends.
I was testing a trend following strategy on 5 minute ES chart in ninjatrader and it came out with a 44% wr without commissions involved.
I am skeptical of how some entry and exists occur on the same bar, also how significant does slippage play a role in backtesting and are there any other limitations I should be concerned about with this backtest?
I am skeptical of how some entry and exists occur on the same bar, also how significant does slippage play a role in backtesting and are there any other limitations I should be concerned about with this backtest?
%%
I dont pay commissions+ dont do es.
ANY thing liquid can be entered in same 5 minute candle sometimes.
AND wonder who gets dividends on that basket of stock derivative ??NOT you
AND that equity curve looks very fake in its smoothness, especially for es;
+ really looks fake for 2009, but anything is possible.5 minutes sounds like noise...…………………………………………………………………………………………………………………….
Now that you mention it it does look too straight, but I set both take profit and stop loss at 5 ticks ($62.5) so it might make sense that dips are not shown because 5 ticks is nothing compared to the cumulative net profit. Any idea why it may look fake?
%%
Because many people do well in a smooth trend;
but a big bear for 2 months + big bull uptrend the rest of the year/2009 is unusual to catch it so well on both sides.
Good points on the long time frame muting peaks+ Valleys.
[ I thought the no commissions most went to 12 months ago may make inverse short term moves much wilder/not on spxs or spxl anyway...……………………………………………………]
Most aim for a smaller risk[loss] than the larger reward[gain] but most anything may work with a hi% win rate
It could very well be that you are taking a lagging signal and getting stopped out by volatility in that back test.
Can't be sure but I would bet that even if you reversed the trades, you would still end up with a loss.
Interesting. The key here is "consistently losing". I remember there is a short mathematical proof in the book Fooled By Technical Analysis by M. Harris (offered in website form only but is worth a read) that a consistently losing strategy is similar to a consistently winning strategy.
Usually losing traders think "what about if I had taken the opposite side...". Obviously for that specific period they would have made money but a bad strategy performance always reverts towards zero before commission and slippage.