Friday, March 5, 2010

Close but no cigar...

It looks like after all of the fight we put up we ended up losing money this week. I traded well today again but caught a loser in the 6S and was slipped a tick ontop of it. But as you can see in the image below... right idea... just a bit early.

So what i'm left with is results in stark contrast to last week. To make sure that i'm not nuts and falling down a primrose path to failure yet again I have analyzed the trades and outcomes from this week and compared them to last week. Two things became very apparent:
  • Missed trades on Monday and Tuesday while I got my feet under me and was having platform issues cost me big.
  • The trades simply weren't there like they were last week.
Missed trades cost me at least 175.00 worth of profit. Not massive, but it makes a difference. I also have decided to move towards giving my trades 1 additional tick of initial risk. I had 2 trades this week where I was low ticked out and could've turned full stopouts into either pars or nice winners. If I can accomplish that with only 1 additional tick of risk I think its worth it.

The more obvious reason for performance (or lack thereof) was the lack of profitable trades. I went back and looked at the actual setups from last week (I take a screenshot of every trade I take and log it) and i'm sorry folks, but they just weren't there this week like last week. There were just awesome trades everywhere that meshed perfectly with what I was doing last week, and that wasn't the case this week for whatever reason.

Take for example the total number of trades taken last week - 34. This week, only 23. Even including the missed trades it would still only be around 27. Conditions simply weren't as great as they were on the prior week.

But I remain committed and optimistic that there is indeed something here so I will continue on next week with the tiny tweak of additional 1 ticks of risk and keep at it.

Certainly not the week I was expecting (it literally went downhill everyday since Monday) but I'm still walking away from it having learned a lot and performed well under the conditions.

Trade Results:
1 Loss
Gross P/L: $-100/per

Weekly Results:
Gross P/L: $55.00/per
Net P/L After Commissions: $-70.35/per
Win/Loss: 54%
Win+Par/Loss: 74%
Profit Factor: .972


Risk Control Master said...

First, congrates on trading real account. It's always different when you betting with real money. I like the way you deal with even losing money on the first week, you have the right attitude. It will lead to success eventually.

One question, how does the "per" mean in your P/L and what's the green/red mean?

Matt said...

Per just meant per contracts traded... Don't want people thinking i'm trading 3 lots and getting these results.

Green/Red just for positive and negative, good bad, and yellow for not great.

Thanks for the kind comment!


TuxedoCat said...

I developed a system a few weeks ago that looked good on backtest results under TradeStation, but then I found out that when you use intelligent routing on TS or don't specifically account for correct ECN routes, the fill times can be horrible. So what should've been an easy 50 cent spread to capture dropped to around 24-30 cents; the profit factor collapsed under real world results.

What you have to realize is that the backtest is just a first step. It gets progressively worse each step of the way, and maybe 1 out of every 20 ideas actually takes off. In a way, we're all just sifting dirt looking for gold in California.

I did the exact same thing as you at 24. My opinion, and this is just my opinion, is that pure-charting techniques without close analysis of order flow or other exogenous variables (fundamentals, news, etc.) just don't work. They look like they work, and look deceptively simple, but it almost never works out that way. I went so far as to try and get a PhD (stopped at masters) in econometrics, and I found that cleared up my vision of the markets.

TuxedoCat said...

Actually, I phrased it wrong. I think you -do- realize the backtest is the first step. I'm just a bit iffy on how you intend to modify your techniques based on a small sample.

I guess if I were you, I -would- be concerned about the missed trades. That's kind of a key thing right there, because maybe the rest of the market does see the same thing you are, but they're quicker to act on the behavior than you are.

Matt said...

Tux -

My backtest results weren't automated in any way. I get what you're saying by not relying on backtests but the trades taken took into account slippage on both entry and exit and aren't really an issue.

The problems this week certainly weren't the cause of order routing.

As far as changing my rules, i'm not changing any of the setups save for the addition of 1 tick more risk. Thats it. Otherwise, its business as usual.

I'm with you - and missed trades is a BIG one. Not only that, probably more so than I touched on there were TONS of trades this week that were overlapped with times I was already in other positions and thus had to pass. Ironically, almost all of these trades were profitable.

And while my "forward testing" from last week made sure to not include these results my overall backtesting results DID assume that I could catch any trade recorded in any market during my trade times, which obviously is not the case with my current account limits.

And obviously this makes a big difference. But for the time being i'm going to restrict myself to one contract only until I see sustained some glimmer of hope at the end of the tunnel and then move from there. A lot of these markets kick off moves simultaneously and it would be nice to catch the same burst in more than one!

Thanks for your input. Its greatly appreciated!