Fully automated futures trading

You expect a Sharpe of 12% unlevered, yes? Perhaps this is obvious, given you've returned nearly 100% on your starting capital.



globalarbtrader said:
Trades

Code: 
Trades take 1

         code contractid     filled_datetime  filledtrade  filledprice
3100      AEX     201504 2015-03-26 13:32:31           -1   482.750000
3098      CAC     201504 2015-03-26 13:26:09           -1  4957.000000
3107      CAC     201504 2015-03-26 16:59:17            1  5008.500000
3102     CORN     201512 2015-03-26 14:06:19            1   418.500000
3105     CORN     201512 2015-03-26 15:06:56           -1   413.500000
3101  EDOLLAR     201809 2015-03-26 14:00:33           -1    97.765000
3099   GAS_US     201506 2015-03-26 13:30:19           -1     2.798000
3095      JPY     201506 2015-03-26 03:06:27            1     0.008392
3104      SMI     201506 2015-03-26 14:23:28           -1  8903.000000
3096      V2X     201505 2015-03-26 13:25:41            6    22.450000
3103      V2X     201505 2015-03-26 14:22:51            1    22.650000
3106    WHEAT     201512 2015-03-26 16:33:50           -1   532.000000


Slippage in GBP, for entire trade


         code  gbpt_slippage_process  gbpt_slippage_bidask  gbpt_slippage_execution  gbpt_slippage_all_trading  gbpt_slippage_total
3099   GAS_US                 -77.08                  3.35                    -6.70                      -3.35               -80.43
3104      SMI                 -45.08                  3.47                    -6.94                      -3.47               -48.55
3102     CORN                 -33.51                  4.19                     0.00                       4.19               -29.32
3105     CORN                  -0.00                  4.19                    -8.38                      -4.19                -4.19
3103      V2X                  -7.31                  3.66                     0.00                       3.66                -3.66
3107      CAC                  -3.66                  1.83                     3.66                       5.48                 1.83
3095      JPY                  16.76                  4.19                    -8.38                      -4.19                12.57
3106    WHEAT                  -0.00                  8.38                     8.38                      16.76                16.76
3098      CAC                  21.94                  1.83                    -0.00                       1.83                23.76
3096      V2X                  21.94                 21.94                     0.00                      21.94                43.87
3101  EDOLLAR                  16.76                  4.19                    33.51                      37.70                54.46
3100      AEX                  76.77                  7.31                   -21.94                     -14.62                62.15

Total slippage: process -12.470000; bidask 68.530000; execution -6.790000; all trading 61.740000; grand total 49.250000

Continuing to cut risk, down to less than £4K a day expected now. Longs in bonds and STIR, and a short in Gas, are the only positions of note now.

LOSS: £9891
£37K below HWM
P&L since inception: £344K

Obviously some chunky losses on equities, but also short crude didn't play out well today. Yesterdays expected risk was £4554, so this was a 2 and a bit sigma day, or roughly one in two months loss. NOT so easy... How to put this into perspective? Since I began writing this thread I've made about 24K, or call it 16K a month. Annualised that would still be a 50% return. My conservative expectation is to make about 12% a year (SR of 0.5).

This is just trading. Unless you're exceptional, or you're running a highly negative skew strategy which one day will blow up in your face, you don't make money every day or every week.

I won't be updating this journal for a couple of weeks as I'm going to be on holiday, although with access to my system so I will be checking things are ticking over, but I want to do the absolute minimum of screen time otherwise.

After that I will probably step back to updating weekly.
More...
 
jj1111 said:
You expect a Sharpe of 12% unlevered, yes? Perhaps this is obvious, given you've returned nearly 100% on your starting capital.
More...

Well Sharpe Ratio is return over risk; so it should not be affected by the amount of leverage you are using.

With a 25% annualised vol target, and an expected Sharpe Ratio of 0.5, my expected return is 12.5% a year with the leverage I've got (I talked about leverage in another post; I think its a poor measure of risk).

Yes I've made a lot more than that this year (350K on 300K starting capital), for two reasons:

- I was running a higher than 25% annualised vol target, and with more capital at risk, for much of the year.
- It's been an exceptionally good year. Even if I'd had the same vol target all I would have made around 57% on my starting capital.
 
Given that it's been such a great year, does that make you nervous that your 25% vol estimate might not be a good estimate (or that, say, its 95% confidence interval might be wider than you think?)

In other words, what would be the probability of earning 150%, given you expected 12.5%?
 
jj1111 said:
Given that it's been such a great year, does that make you nervous that your 25% vol estimate might not be a good estimate (or that, say, its 95% confidence interval might be wider than you think?)

In other words, what would be the probability of earning 150%, given you expected 12.5%?
More...

To be precise I target 25% annualised daily vol. So its the daily vol I'm targeting, not the annual vol. Because of time series correlation, the two might not be the same. Also I get a new figure for daily vol every day, so its easier to know if its in line. I watch my daily vol like a hawk. My vol estimates are pretty good. I've never had a 3 sigma day, up or down. I'm very comfortable with that.

A better test is to say what is the probability of making 56%, given a 25% sigma, since that abstracts away from the change in my risk target (to repeat 56% is what I would have made if I hadn't adjusted my risk target throughout the year). The answer depends on your underlying Sharpe Ratio, which is unknown. I use 0.5 as a conservative figure. But the backtest, using robust out of sample fitting, comes in at 0.9 Sharpe.

If the true Sharpe Ratio is 0.5, then the chance of making 56% is around 4%.

If the true Sharpe is 0.9, then the chance of making 56% is 10%.

So what we have is eithier a one in 25 year return, or a one in 10 year return.

(This assumes a normal distribution, which for annual returns is probably okay).

So my true Sharpe is probably higher than 0.5, but I'm comfortable with this as a conservative lower bound.
 
softdown said:
Hello globalarbtrader,

When you have the time, could you talk a bit about your back-testing methodology ? One interesting topic for me is the relation between in-sample and out-of-sample results.
More...

This is probably something I will put on my blog in due course, and I'll post a link here.
 
globalarbtrader said:
To be precise I target 25% annualised daily vol. So its the daily vol I'm targeting, not the annual vol. Because of time series correlation, the two might not be the same. Also I get a new figure for daily vol every day, so its easier to know if its in line. I watch my daily vol like a hawk. My vol estimates are pretty good. I've never had a 3 sigma day, up or down. I'm very comfortable with that.

A better test is to say what is the probability of making 56%, given a 25% sigma, since that abstracts away from the change in my risk target (to repeat 56% is what I would have made if I hadn't adjusted my risk target throughout the year). The answer depends on your underlying Sharpe Ratio, which is unknown. I use 0.5 as a conservative figure. But the backtest, using robust out of sample fitting, comes in at 0.9 Sharpe.

If the true Sharpe Ratio is 0.5, then the chance of making 56% is around 4%.

If the true Sharpe is 0.9, then the chance of making 56% is 10%.

So what we have is eithier a one in 25 year return, or a one in 10 year return.

(This assumes a normal distribution, which for annual returns is probably okay).

So my true Sharpe is probably higher than 0.5, but I'm comfortable with this as a conservative lower bound.
More...


Thanks for the explanation.
 
globalarbtrader said:
Review of first years performance
But a downside of that is when (not if) I start to lose money, and enter say a 15 or 25% drawdown, it will be harder not to be patient since if I'm not careful I'll have unrealistic expectations. Anyone starting to trade this kind of system in 2009 would have really been hacked off by March 2014. If they'd thrown in the towel then they'd have missed out on the incredible performance we've seen since then.
More...

Doesn't this provoke an interesting question. Is this risk profile really suitable for an individual?
I can see a fund comprised of long term investments of a small % of liquid net worth tolerating multi year drawdowns or lackluster performance on a strategy which ought to outperform in the long run. Yet "in the long run, we're all dead."

I think a strategy which either doesn't trade frequently enough or doesn't capture enough advantage per trade to recover to new equity highs quickly is very hard for an individual to cope with while those who can afford to take a longer term view may find it more valuable.

Would be interested in your thoughts on this.

I have always viewed trading as being on a continuum from insolvency to retirement. You haven't made it until you've taken your chips off the table for the last time. Although there are always some who are clearly a lot closer to making it than others. Have heard too many stories of blow ups, suicides, etc even amongst multi decade veterans. Flash crashes, errors in judgment, vice, human nature, complacency, or just old man vig.

The successful traders I know all seem to have an undesirably high emotional exposure to their P&L - which persists well beyond the period of any initial doubts about whether they can make it in this business. There is the eternal trade off between having too much capital exposed - or the equally dangerous having too little exposed which then forces you to spend more time in the markets as you didn't capitalise on the opportunities available. It seems people are sensitive to risks to capital but not appropriately sensitive to risks to time. I would have liked to have realised this earlier in life.

Speaking personally I have found a generalised anxiety arising in periods of poor strategy performance. It can affect how one views their overall competence and must be guarded against. In later years I determined that it was preferable to turn off the more marginal but still +EV strategies and keep the strategies which make higher and more consistent returns per contract traded. Somebody else can have the more marginal stuff because the cost in anxiety is not worth the increase in performance.* So one can end up with a strategy which trades less but is very very consistent.

For you, it appears you are capturing an effect which requires long duration of market exposure and therefore this is not an option for you. Thanks for your contributions, always interesting to see how others are doing things.

* - yet in making that trade off I am again exposed to more years in the market to achieve a goal - but otherwise "the game isn't worth the candle"
 
You must log in or register to reply here.
Back
Top