Mystery Software Journal

The signal on the USZ8 contract switched from SHORT to LONG tonight after the close, so I exited the short (and went long) at 117 - 28.5. That's a loss of 4/32nds, or .125 point, which is -$125 in this case.

The current status is as follows:

Open Positions:

Naz100 mini (NQU8) - LONG @ 1,941.75
30yr bond (USZ8) - LONG @ 117 - 28.5
Yen (JYU8) - LONG @ $.9151
S&P mini (ESU8) - LONG @ $1,267.50

Closed equity: - $1,729

Open equity: -$977

22 down, 8 to go.
 
Only one signal tonight: OUT of the long bond contract. I sold at 117 -29, and since I went long at 117 -28.5, once again we get to record a one-half of 1/32 of a point gain, which is $15.62. I was surprised when this happened the first time, it's even more surprising to me to see it happen again -- but there it is.

The current status is as follows:

Open Positions:

Naz100 mini (NQU8) - LONG @ 1,941.75
Yen (JYU8) - LONG @ $.9151
S&P mini (ESU8) - LONG @ $1,267.50

Closed equity: - $1,713

Open equity: -$205

23 down, 7 to go.
 
Oops, I was going too fast earlier today and missed a signal. I had an OUT signal for the ES contract, which I executed a short while ago at $1,279.25. That represents an 11.75 point gain, or +$588. (It would have been a larger gain if I hadn't been careless, as the value has dropped after hours)

The current status is as follows:

Open Positions:

Naz100 mini (NQU8) - LONG @ 1,941.75
Yen (JYU8) - LONG @ $.9151

Closed equity: - $1,125

Open equity: -$935

24 down, 6 to go.
 
Quote from chanelops:

Oops, I was going too fast earlier today and missed a signal. I had an OUT signal for the ES contract, which I executed a short while ago at $1,279.25. That represents an 11.75 point gain, or +$588. (It would have been a larger gain if I hadn't been careless, as the value has dropped after hours)

The current status is as follows:

Open Positions:

Naz100 mini (NQU8) - LONG @ 1,941.75
Yen (JYU8) - LONG @ $.9151

Closed equity: - $1,125

Open equity: -$935

24 down, 6 to go.
More...

You're still long NQ at 1941.75? The current quote I have is 1898.75. Way to hold on to a loser. That's a huge drawdown to take in a system. Overall, from reading your first page, then the last page, you ought to drop this mysterious black box, and actually learn how to trade.
 
Had several signals last night that I acted on. First, an NQ signal to exit that trade, which I did at 1,905, for a loss of 36.75 points, or -$735. Also got a signal to exit the yen trade, which I did at $.9140, for a loss of .11 points, or -$137.50.

Then two new positions: LONG on the US contract, executed at 117 - 30.5, and SHORT on the ES, filled at $1,296.

The current status is as follows:

Open Positions:

30-yr bond (USZ8) - LONG @ 117 - 30.5
S&P mini (ESU8) - SHORT @ $1,296.00

Closed equity: - $1,998

Open equity: -$203

26 down, 4 to go.

P.S. To the last poster: if you'd read the full journal, you'd see that this is an attempt to validate trading the signals from this software without any sort of money management, taking only entries and exits as indicated. Obviously, if I was using stops and variable position sizing, the results would be very different.
 
Got a signal tonight to cover the ES short, which I did at $1,284.50, for a gain of +$575.

The current status is as follows:

Open Positions:

30-yr bond (USZ8) - LONG @ 117 - 30.5

Closed equity: - $1,423

Open equity: -$641

27 down, 3 to go.
 
Got an OUT signal on the Tbond contract, and exited at 118 - 0.5, for a gain of 2/32nds, or +$63.

The current status is as follows:

Open Positions:

None

Closed equity: - $1,360

Open equity: 0

28 down, 2 to go.
 
I assume you chose 30 trades because that is the number considered to be the minimum size for a good sample.

I've found that different instruments trade differently. I'm guessing that you backtested each of the instruments independently. Is this the case? If so, I think you should continue posting until each of your instruments has traded 30 times. That would probably be a fairer test of the "Mystery Software."

Dropping gold as you have done may not be a fair test as you are skewing things away from failure. However, a lot depends on how long your backtest was run and what the market looked like when you developed and tested.

Still, it's your time and money, so thanks for your posts.
 
Whimsy,

Thanks for your helpful comments. You are right as to why I chose 30 trades. And I think you are also correct that it would be fairer to continue posting until each instrument has traded 30 times. I also agree with you that dropping gold skews the validity of the testing. On the other hand, I wasn't trading gold any more, and I wasn't going to continue doing it just to be able to post. And I didn't want to post theoretical fills, although that might have been OK.

The way things stand right now, I'm leaning towards concluding that the software does not work good enough by itself to be able to be used on a standalone basis, without a money-management layer. At a minimum, that layer would include stops. I think that with some intelligent stops, these results would have been vastly different (better). But since stops almost invariably reduce your total profit, it would have been nice if it had turned out they could have been avoided, except for maybe some sort of catastrophe stop.

If I conclude that money management is required, then I may do this whole test again from scratch with the addition of some MM rules, and incorporate some of your good suggestions.

Thanks for the post.
 
New signal tonight: LONG on the ES contract, filled at 1,273.50.

The current status is as follows:

Open Positions:

S&P mini (ESU8) - LONG @ $1,273.50

Closed equity: - $1,360

Open equity: 0

28 down, 2 to go.
 
You must log in or register to reply here.
Back
Top