Kudos to MMs

GTS · Sep 23, 2010

Quote from Pekelo:

No, 1 point here or there doesn't make that much of difference. I question your P/L column, that should show close to +5 and -9. (instead of -21 and -12)
I guess you are trying to show the theoretical difference, but who cares once a trade is closed?

My P/L column is based on a keeping an average position (basis) for all trades, so after 4 short trades ( 1089, 1104, 1107, 1121.75) the overall short position basis was 1105 - if you close out 2 units @ 1116.75 the loss is ~22 (1116.75-1105 x 2)

Nitro appears to be tracking each trade separately and not averaging them together - nothing wrong with that but as a futures trader that's not how I'm used to thinking so I am tracking it my way. Mathematically both methods will end up netting out to the same overall gain/loss in the end (when we are flat)

Code:

2-Sep	 Short 	-1	 1,089.00 	-1	 1,089.00 	
3-Sep	 Short 	-1	 1,104.00 	-2	 1,096.50 	
9-Sep	 Short 	-1	 1,107.00 	-3	 1,100.00 	
13-Sep	 Short 	-1	 1,121.75 	-4	 1,105.44 	
15-Sep	 Cover 	 2	 1,116.75 	-2	 1,105.44 	 (22.63)
16-Sep	 Cover 	 1	 1,118.00 	-1	 1,105.44 	 (12.56)
17-Sep	 Short 	-1	 1,133.00 	-2	 1,119.22 	
20-Sep	 Short 	-1	 1,142.50 	-3	 1,126.98 	
23-Sep	 Cover 	 1	 1,124.00 	-2	 1,126.98 	 2.98

Pekelo · Sep 23, 2010

Quote from GTS:

Mathematically both methods will end up netting out to the same overall gain/loss in the end (when we are flat)

Oh I see. I think you are complicating the accounting because after each close out you have to recalculate the average. There might be more add-ons what will also change the average...

Although I did mention earlier that it is still nice to know where the average of the position is compared to the market, but I wouldn't use it for accounting purposes for the above mentioned reasons.....

nitro · Sep 23, 2010

Quote from johnnyqpublic:

Nifty! Is the trading automated?

Not initially. It is no big deal since this system holds positions for extended periods, and exact entry and exit is noise that averages out over time.

GTS · Sep 23, 2010

Quote from Pekelo:

Oh I see. I think you are complicating the accounting because after each close out you have to recalculate the average. There might be more add-ons what will also change the average...

Closing out a position does not change your average (basis) only adding additional positions would. I threw it into Excel so keeping the running basis is automatic.

I find Nitro's method of referring to each position by the edge that existed during entry hard to follow (e.g. "We are forced to take profit on the 35 units") and it makes it difficult to follow the current overall position profit/loss.

Again to each his own - if Nitro doesn't want me posting it I'll keep it to myself...I think the current method of posting entries/exits makes it very difficult for casual visitors to the thread to know what is going on.

johnnyqpublic · Sep 23, 2010

Quote from nitro:

Not initially. It is no big deal since this system holds positions for extended periods, and exact entry and exit is noise that averages out over time.

I see. I remembered you stating once that you were spending quite a bit of time trying to automate trading based on NFV, and wondered if this might be a sort of test run.

nitro · Sep 23, 2010

Quote from johnnyqpublic:

I see. I remembered you stating once that you were spending quite a bit of time trying to automate trading based on NFV, and wondered if this might be a sort of test run.

Well, I have ideas for turning it HF, but that is a wish and not realistic right now.

nitro · Sep 23, 2010

Quote from GTS:

Closing out a position does not change your average (basis) only adding additional positions would. I threw it into Excel so keeping the running basis is automatic.

I find Nitro's method of referring to each position by the edge that existed during entry hard to follow (e.g. "We are forced to take profit on the 35 units") and it makes it difficult to follow the current overall position profit/loss.

Again to each his own - if Nitro doesn't want me posting it I'll keep it to myself...I think the current method of posting entries/exits makes it very difficult for casual visitors to the thread to know what is going on.

I do it that way only because that is the way I think about it. It might be better to post a spreadsheet with positions, PnL, etc.

I don't mind your posts at all - they are constructive.

nitro · Sep 23, 2010

NFV 1106.13. SPX 1124.83.

Nothing to do. We need to see convergence to 14 to act in reducing units, or going back to 35 divergence to add units. If you want to act overnight and ESZ0 is lower by > 4 handles (edge would be about 14 then if NFV stayed still) to take profits on your 24 units, I can't blame you.

Beware though, if NDX was not green today, SPX would have been blasted down 17.

nitro · Sep 23, 2010

GTS, this looks right (except that action was taken today which the table does not represent). The only difference is the way we think about it. I don't keep an average basis. In fact, the way brokers do it, they take the worst position when you exit and that is the one you covered. So for example, in the +2, those would have matched the 1089 and the 1104 trades, not the way I think about it as the units which forced me to exit. Brokers force you to take losses first, not profits.

You are right that the way I post it doesn't maintain state very well so that someone trying to follow this would have to recreate from posts. Other than every few pages posting a table like you did could I improve the presentation. I will try to do better....

Code:
Quote from GTS:

This is a summary I have since 9/2 for those that like to track using an average basis rather than treating each position as a discrete unit.

Nitro, if I made any mistakes please let me know.

Code:

Date Action Quant Price Position Basis Realized Gain/Loss 09/02/10 Short -1 1,089.00 -1 1,089.00 09/03/10 Short -1 1,104.00 -2 1,096.50 09/09/10 Short -1 1,107.00 -3 1,100.00 09/13/10 Short -1 1,121.75 -4 1,105.44 09/15/10 Cover +2 1,116.00 -2 1,105.44 (21.13) 09/16/10 Cover +1 1,118.00 -1 1,105.44 (12.56) 09/17/10 Short -1 1,133.00 -2 1,119.22 09/20/10 Short -1 1,142.50 -3 1,126.98

nltro · Sep 24, 2010

Quote from nitro:

NFV 1106.13. SPX 1124.83.

Nothing to do. We need to see convergence to 14 to act in reducing units, or going back to 35 divergence to add units. If you want to act overnight and ESZ0 is lower by > 4 handles (edge would be about 14 then if NFV stayed still) to take profits on your 24 units, I can't blame you.

Beware though, if NDX was not green today, SPX would have been blasted down 17.

With NDX up another 24 handles premarket and SPUZ up13.50, how much anti-blasting are you willing to absorb? I guess that no matter how many add points you have, there is a 100% chance over time that there would be sufficient convergence of SPX/NFV that will cause ALL trades to be closed. What you fail to incorporate in your RTM model is the possibility of a 3+ std dev move that does not retrace, ie TREND. What you are doing is akin to picking up nickles in front of a steamroller. 90% of the time you will get lucky. The time you fail you give it all back x2.