Rol's Trading Journal

[heech]

Quote from Rol:

If you have read this journal, you would know that my "purpose" is to try to match the back testing results in real world trading. If you want to nit pick over individual trades then perhaps you are missing the purpose. I am as incredulous as anyone that the system can achieve the back testing results, which is why I chose to document this experiment. I have not yet been able to prove that it does not make money. Take it anyway you want.

You have to understand if the equity curve is purely showing "realized results", I'm not at all incredulous as to its shape.

There's another very hot thread on this site right now, the martingale discussion. If you were to look at the equity curve (both hypothetical and real) for a martingale on a "realized" basis, it would be far more impressive than what you've generated. This is precisely why marked to market numbers are very important.

In any case... its your show, and as I said, I'm not making any demands. I subscribed to this thread because I thought the results were exciting. But they simply aren't exciting if they're not marked to market.

EDIT: Thread to Martingale discussion:

http://www.elitetrader.com/vb/showthread.php?threadid=215965

 

[DGunz]

Heech,

You bring out a good point, for the sake of having accurate stats and charts to evaluate a model, 3 days average hold time for not recording unrealized loss/profit wont materially change all stats that he has recorded of the past year.

I agree that m2m on the stats would be a better gauge, the longer the hold time for unrealized losses the more unrepresentative the equity curve. However if the average loss hold days was 1 day flat.... then I would say that the curve is still highly representive because the losses are shortly recorded.

Rol, if either presentation doesnt matter to you, then why not try and put out m2m results. It wont change your stats you recorded on a closed basis anyway and it may help you see something you didnt see before.

 

[Rol]

Quote from heech:

You have to understand if the equity curve is purely showing "realized results", I'm not at all incredulous as to its shape.

There's another very hot thread on this site right now, the martingale discussion. If you were to look at the equity curve (both hypothetical and real) for a martingale on a "realized" basis, it would be far more impressive than what you've generated. This is precisely why marked to market numbers are very important.

In any case... its your show, and as I said, I'm not making any demands. I subscribed to this thread because I thought the results were exciting. But they simply aren't exciting if they're not marked to market.

EDIT: Thread to Martingale discussion:

http://www.elitetrader.com/vb/showthread.php?threadid=215965

I didn't think mark-to-market was that big of a deal. My strategy is not martingale style. I dont average down to zero, by doubling my bet after getting into a loser. In fact my strat makes no distinction regarding position size whether it is currently a winner or loser. I am diversified across many stocks. The entire market would have to go to zero and stay there forever for my system to crash. If the market does that, then maybe money won't matter anyway. At any rate, from now on I will post marked-to-market equity curves instead of only realized profit loss for those who care. Here is a marked-to-market curve since going live with my system starting 9-15-10.

TradeStation TradeManager Analysis Settings

Start Date 9/15/2010
End Date 2/25/2011

Portfolio Template Mark to Market

Initial Capital $30,000.00

 

[heech]

Quote from DGunz:

I agree that m2m on the stats would be a better gauge, the longer the hold time for unrealized losses the more unrepresentative the equity curve. However if the average loss hold days was 1 day flat.... then I would say that the curve is still highly representive because the losses are shortly recorded.

I'd definitely agree that marked to end of day prices, or even end of week prices would be representative/meaningful, if done over a long enough timeframe. The important thing is that these samples are consistent across the entire portfolio.

But if these prices are cherry-picked... winning stocks are priced at market prices when they're profitable, losing stocks aren't priced unless they're profitable... then what conclusions could we possibly draw from the fictional graph that results?

 

[shortie]

a daily chart would make more sense to look at. for example, when did ~#800 trade drawdown occur? i have no idea, it does not relate to me in any way. if i knew that it happened in december, it would make more interesting for me to think about the market at that time, etc.

 

[DGunz]

Rol,

Whats your average hold for a unrealized loss? As long as the amount of days isnt something like 4 months, it is probably likely that the shape of the curve is basically intact.

Is it easy for you to show m2m, if it is, perhaps you could just leave it that way. Personally I am partial to knowing the m2m rather than closed, but its your journal.

 

[DGunz]

Quote from heech:


But if these prices are cherry-picked... winning stocks are priced at market prices when they're profitable, losing stocks aren't priced unless they're profitable... then what conclusions could we possibly draw from the fictional graph that results?

I give him props for NOT cherry picking because he is recording closed trades, it is consistant in that he sticks to that metric. I dont think hes recording only the winners at market and not the lossers. I think the losers are usually the ones that end up being held longer for unrealized losses, theres always a slight unrealized loss somewhere.

 

[Rol]

Total Net Profit $1,695.17
(Per Share) $0.06
Gross Profit $7,720.92
Gross Loss ($6,025.75)
Profit Factor 1.28
Total Number of Trades 219
Percent Profitable 64.38%
Winning Trades 141
Losing Trades 78
Avg. Trade Net Profit $7.74
Avg. Winning Trade $54.76
Avg. Losing Trade ($77.25)
Largest Winning Trade $232.51
Largest Losing Trade ($369.89)
Max. Consecutive Winning Trades 26
Max. Consecutive Losing Trades 15
Total Shares/Contracts Held 28502
Total Commission $533.40
Return on Initial Capital 3.26%
Annual Rate of Return 41.85%
Buy & Hold Return -0.61%
Trading Period 28 Dys
Percent of Time in the Market 60.66%


Well it was not an easy week, mostly spent just trying to get back to b/e. I view this as a consolidation period for my equity curve. Profits should take off again eventually. When I look at the curve, I can almost see areas of support and resistance, as if it were a stock trading. Naturally, the 10% intraday DD was pushing my comfort zone, so I have scaled back on the max number of open positions to keep it closer to my goal of 8% max DD. While even 8% may sound like a lot, I might never trade at all if I constantly worry about a black swan event. I just need to be able to hunker down in my foxhole, waiting for the dust to settle, and come out fighting another day. It would not be an account blow up, but rather just pulling back to a previous "support level" on my equity curve. Even so, it is better than buy and hold, IMHO.

The PF is less than my goal of 2 since starting 1-31-10, but I expect it to average back up longer term. PP is still close to my goal of 66%. Buy and Hold return for the period is negative, so even mark-to-market, my strat is beating buy and hold. I also think I will still meet my goal of not having a negative month.

 

[Rol]

Quote from DGunz:

Heech,

You bring out a good point, for the sake of having accurate stats and charts to evaluate a model, 3 days average hold time for not recording unrealized loss/profit wont materially change all stats that he has recorded of the past year.

I agree that m2m on the stats would be a better gauge, the longer the hold time for unrealized losses the more unrepresentative the equity curve. However if the average loss hold days was 1 day flat.... then I would say that the curve is still highly representive because the losses are shortly recorded.

Rol, if either presentation doesnt matter to you, then why not try and put out m2m results. It wont change your stats you recorded on a closed basis anyway and it may help you see something you didnt see before.

DGunz, I already beat you to the idea of displaying mtm before I read your reply . It will show better the "sweat" one has to endure, but I think I have already been discussing that from the beginning. Now even more with this journal, you will see the good, the bad, and the ugly. Disclaimer - Read at your own risk :eek:

 

[Rol]

Quote from DGunz:

Rol,

Whats your average hold for a unrealized loss? As long as the amount of days isnt something like 4 months, it is probably likely that the shape of the curve is basically intact.

Is it easy for you to show m2m, if it is, perhaps you could just leave it that way. Personally I am partial to knowing the m2m rather than closed, but its your journal.

DGunz, as you can see in the MTM chart below average hold time for winners is 1 day, 19 hours, and 15 minutes. For losers it is 4 days, 2 hours, and 12 minutes. As I said on my very first post, my average hold time is 2 days. I admit my style is foreign to most people, and I am hoping I can use that to my advantage.