Do my numbers make sense?

[Adim]

If you are only testing against bull market and only going long just BTFD already and be done with it.

Just don't think that is what trading is.

This is a Grasshoper test, yes? Did you see the part where I showed how the program did during the recent correction?

On another note... is it a test of a good trading program that the manager of the program is not allowed to read newspapers to learn about a rapidly spreading epidemic in China which may not yet be reflected in the SP500?

I have come here to learn something, not to be dumped upon, but thank you for testing my manhood

 

[lindq]

sorry. but true. i was not in the market from December 2007 until January 2009. i come from EM debt background, i am good at sitting on my hands. I also went to cash in January 2020 because I read Chinese and could see the news. if this is incredible, then I am really worried.

if maximum adverse excursion means "worst trade" it is was an embarassing -1.10%

You report that your average holding time is 13 days and you average 33 open positions, and trade long only.

Hopefully you have serious risk controls in place. You'll need them.

 

[Adim]

> You report that your average holding time is 13 days and you average 33 open positions, and trade long only.

i am not very leveraged. in fact, most days i am not even 100% invested

 

[Adim]

I dont know guys... I was NOT looking for sniping. I already know from my wife that i am mentally retarded, I hardly need you to help me with that insight.

Does anyone have anything useful to say?

 

[deaddog]

By 9/30 SPY was down -5.3% and i was down -2.63%. Sharpe Ratio seems to be something stupid like 16+.

where am i going wrong?

What are your sept stats?
Winners, losers, etc. ?
What was your biggest loser, winner, win %, R to R?

Didn't your backtest have any losing months?

 

[hilmy83]

Can't be both.

Lolol almost spit my coffee

 

[traider]

Yes... thank you. i have thought about this. but... the reason why I did not test it through the bear markets is three fold:

1) first, i expect the immediate future to be more bull than bear, so bull market testing is more appropriate for the present (i think). i assume no interest rate hikes until end 2022 (perhaps middle 2023). (manual trader talking)

2) in the case of both bear markets GFC (2007-2009) and Covid (2020) i could see miles off to get out of the market ahead of time. both crashes were amply advertised ahead of time for anyone that actually paid attention. (I went to cash Dec 2007 and again January 2020). So, excepting an event like 1987 or 2001 (9/11), in which case there isn't a whole lot anyone can do except pull the plug, i am not too worried about a bear market performance. i assume i will just go to cash. yes?

3) because the routine uses a kind of price discovery gizmo, it can't really be back tested. all that can be backtested are the general conditions under which as stock should be given to my routine to trade. its the actual trading which gives me the clue -- in this case, September hiccup looks very good relative to market in general. but I will backtest the starting conditions like you suggest going back to 1998. i will take out the periods when I was in cash -- if I knew it then, I would not have been trading then, yes? bad thinking on my part?

It feels like you are not calculating sharpe correctly , you should double check. You outperformed SPY but the magnitude seems to be more like 2x or 3x and not what your sharpe says. HFT turnover gets you to sharpe 10+ but it is rare for other type of strategies.
How did you manage to see the decline coming for Covid or GFC? Do you have any false positives for detecting market crashes because false positives are also costly.

 

[Adim]

traider
it is entirely possible that i am not calculating Sharpe correctly. would you like to help me with that? is there some way we can text directly?
also, i just learned there was something called Sortino. Not from qlaid (or whatever his handle sez), 'koz of course his job is to haze newbies.
like i said, I have been a manual trader all my life, I am just learning this stuff.
Avoiding the Covid crash was not hard. All you had to do is read the press. It helped that I read Chinese and Japanese, and had lived in Asia, and had been in Malaysia during the SARS epidemic. beyond that it was just a matter of reading the press and drawing on your own experience. I was actually short oil into negative prices, as a result. i gather from the few replies here that quant trading means one must be too proud to know anything beyond the numbers in front of him. (is this weird?). "we dont read the press, we dont interpret the news"? is that, like, virtuous, or something?
GFC was perhaps less obvious, but i was not computer trading at the time, all manual, and through an European bank with extremely high fees (like $100 per trade minimum, i am not making this up, this is how the rest of us -- the other half -- live). and, as a result, sometime in 2005 i went all cash in AUD at .78c because I could see nothing else that was worth owning. at that time AUD paid 7.5% per year. with market valuations at what they were, it just seemed smart to own an undervalued currency (on burgernomics basis) paying 7.5%. i started dipping my toe in by buying MBIA CHF 18 month debt in december 2007 at 70c. (it eventually went down to .65c in January 2009).

 

[SunTrader]

This is a Grasshoper test, yes? Did you see the part where I showed how the program did during the recent correction?

On another note... is it a test of a good trading program that the manager of the program is not allowed to read newspapers to learn about a rapidly spreading epidemic in China which may not yet be reflected in the SP500?

I have come here to learn something, not to be dumped upon, but thank you for testing my manhood

You can talk manhood, I'll stick with trading. Going loooong AND short, no pun intended. LOL

 

[Adim]

deaddog

What are your sept stats?
Winners, losers, etc. ?
What was your biggest loser, winner, win %, R to R?

Didn't your backtest have any losing months?

i cant say if my backtest had any negative months because i did not run it as a backtest of a portfolio (maybe i should, Im just a newbie, should i? i just assumed that if you get every trade right, then portfolio will follow?). so... i ran it as a backtest of so many trades, without agglomeration. the backtest showed various rates for winners/losers depending on sector. so, for example, SP500 individual stocks showed 68/32. but if you tested for sector ETFS (in the XL..., family), then it was more like 72/28.

your other questions, i can only answer regarding my actual trading results (since April 2, 7.5 months): my win ratio was 93% (?! seems really weird). my biggest winner was 12.78% (a real outlier, the next one was 10.19%, the next one after that 8.06%, then 6.44%, then everything lese 4-handle or less) my biggest loser was -1.1%. what is R to R?