Fully automated futures trading

Kernfusion said:
So what logic is currently included into the automatic rule-weight calculation, apart from dropping expensive rules?

In the Strategy report "Forecast weights" I can see that e.g. SP500_micro (a cheap to trade instrument) has breakout20=1.2 but breakout160=NaN, so a more expensive variation is included but a cheaper one is dropped. Are you also incorporating the historical SR as described here (or a simpler version as in the spreadsheet here) or even doing a full mean-variance optimization with some robust bootstrapping to fit the rules daily?
And also, as I understood from the blog post, these forecast weight estimations are done by pooling all instruments, so the same forecast should mostly have the same weight for all instruments (unless some expensive rules were excluded for a particular instrument, which causes re-normalization of the remaining rule-weights for this instrument), but in this report the weights are all over the place from instrument to instrument..
More...


So this time, just for kicks, I fitted everything on a per asset class basis. One of the main reasons is that my fitting just broke with all the instruments present, my laptop literally fell over. And from here, I know that asset class fitting (ok there I used clustered groups but close enough) is as near as dammnit

And here are the parameters I used:

Code: 
    system.config.use_forecast_weight_estimates = True
    system.config.forecast_post_ceiling_cost_SR = 0.13
    system.config.forecast_weight_estimate['pool_gross_returns'] = True
    system.config.forecast_weight_estimate['ceiling_cost_SR'] = 9999
    system.config.forecast_weight_estimate['date_method'] = 'in_sample' ## because production
    system.config.forecast_weight_estimate['equalise_SR'] = False

That means that expensive rules aren't just ruled out at the end, but will also have a lower SR in the optimisation. I'm using my historical SR incorporation (the complicated way), but it's not a full MVO because that would produce bad results. This results in weights that can be quite different within an asset class.

I also removed all weights less than 1%, which is why the weight matrix is so sparse eg for Bitcoin.

I've become fairly blase about this because I just have so many instruments, and so many trading rules most of which are highly correlated, that almost any vaguely sensible fitting process will produce almost the same outcome. Perhaps this is a case of do as I say, not do as I do....?


Rob
 
globalarbtrader said:
A quick update (more will follow in my usual annual review blog post in a few weeks time), good performance recently - now at a HWM and YTD about 12%. But don't worry, things will get worse, for two reasons: firstly the usual universe punishing my hubris for posting at a HWM, but also because I've substantially increased the allocation to my futures account.

I've also decided to no longer reveal exactly what that cash allocation is. In the past I've been quite open about the exact amount but then had to hide what my allocation to futures is as I don't really want people to infer my exact net worth by dividing the two numbers out (though you can probably have a guess, but I will say the very small number of people who do know my net worth are surprised how poor I am relatively speaking - certainly much poorer than if I had worked for another ten years in the hedge fund industry!). Instead I think it makes more sense to be open about what my futures risk allocation is but not the $ amounts.

So FWIW my risk allocation now is around 40% in futures, 5% in bonds and 55% in equities (cash allocation to bonds is obviously higher, and may well be higher anyway since I hold a chunk of very short term bond ETFs in my futures account which I label as cash).

[Of course given my allocation has increased; and I was open about my account size before, this does give you a lower bound on my net worth, but I leave that calculation as an exercise for the reader who really cares about this sort of thing - and people do google my net worth(!)]

The last real money figure I will ever share with you then is this; my accumulated profits in this account over what will be ten years of trading at the end of this month now come to over $1 million. Which means I can now write a click-baity.book called million dollar futures trader or something. Of course I probably saw gains/losses of a million bucks every few seconds back in the day, so forgive me for not being that impressed with myself.

Anyway back to the allocation increase. Reasons for this are several;

  • I had a large money losing long only investment which allowed me to take a big tax loss.
  • I put a bet on long duration late last year (buying UK and US long maturity government bond ETFs) which was very good in terms of timing getting in, but I was a bit slow getting out and gave quite a bit of it back. If I had done this trade properly I would have bought long dated bond futures and had a proper stop loss; instead I entered into it without any clear exit plan, violating all my own rules. Anyway as it was I did decide to close this overweight offsetting the relatively modest profits against the tax loss, and leaving me with a chunk of change to reinvest.
  • I recently added a huge chunk of instruments to my futures strategy, though I do need to check I can actually trade them all (see recent post on issues with WhatIf orders), and it seemed a good idea to also pour more money in so I could take better advantage of this.
  • Looking back at my past performance, I seem to be best at adding alpha where I'm actively trading - futures and also my UK stock portfolio; in contrast I haven't done terribly well at long only beta allocation in stock/bond ETFs (although we can argue about the statistical significance of these figures).
  • It also seems more plausible than ever that the pure sharpe ratio for a massively diversfied futures strategy will be higher than a long beta strategy in equities and bonds; and even if it wasn't then all the work I have ever done suggests an allocation to trend following should be much higher than I've ever had it before, purely on a diversification basis. And my strategy isn't just trend following.
  • In the past I've used this trading account as a 'piggy bank', only withdrawing cash above the HWM and not compounding my account size above that. Instead I relied more on long only investment dividends for income. I'm happier now to allow the capital at risk to fluctuate with the account size, compounding profits when I can, but also withdrawing money if I need it regardless of whether I am profitable or not - basically how I treat the rest of my portfolio.
  • Having now traded for 10 years independently I'm now fairly confident that my coding and strategy design is robust enough that I'm not going to lose all my money through some huge error; sure the risk premia may go away for the next 10 years but then it could also go away for equities and bonds.
  • There is a potential risk in having a large balance at a single broker, but as has been discussed on this board before it's very small. If IB go down, then I suspect we will all have bigger problems to worry about.
  • There is a potential loss in tax efficiency, since I can time/manage my ETF trades, just as I alluded to above. But with dynamic optimisation it's perfect feasible to do some manual early closing on positions just before tax year end to realise a loss or gain where that would be preferable to waiting a few days (and if they are reopened in the new tax year, that's fine, as long as they are rolled at the same time then they will be treated as new instruments for tax purposes).

None of this adds up to a quantitative justification, but I think it makes sense.

And I'll probably let the account balance creep up some more, if possilble, which is more likely if I can take further long only tax losses and see profits in the futures account.

Anyway hope everyone else is doing okay

Rob
More...

why aren’t you allocating more money to your strategy vs passive investments in mainly equities?
 
newwurldmn said:
i meant more than 50%. like why not 80%.
More...

OK to be clear it's currently 40% and used to be significantly less. I've covered some of these reasons before, but:

  • Having 80% of my wealth in a leveraged futures account with a single broker controlled by software written by a complete amateur developer feels risky. It's taken me 10 years to trust this combo enough to move up to 40%.
  • A good proportion of my wealth is in tax shelters, which in the UK at least are difficult if not impossible to use to trade futures. Although some of this can be offset by the higher risk target in my futures account, and therefore better tax efficiency, I would struggle to get near 60% even if I put all my non sheltered investment money into the account. Since I fill my tax shelters to the maximum every year, this constraint will bind more as time goes on.
  • Although I'd like to get higher than 40%, the speed of this will be limited by how quickly I can liquidate long only investments without incurring capital gains tax. Getting to 40% was the most I could do this tax year.
  • As I alluded to in the original post systematic futures trading is quite tax inefficient compared to long term holding of investments where I can time the sale. However the UK capital gains tax free allowance has been reduced to just £3k as of April, so this benefit is worth much less than it was.
  • My equity/bond holdings are long only, but not passive; I do rebalance them (slowly) using simple simple momentum and value rules; so my risk allocation to eg momentum is higher than you think.
  • More of a problem than before, but I preferred the steady drip of long only dividends as an income stream. However in the last 10 years things have changed, my non investment income is higher (it started at zero), and with more of my savings in tax shelters I was receiving much less in dividends anyway; plus with higher interest rates my IB account pays a fair chunk of interest even if the futures portfolio isn't performing. Nevertheless, I know I will still find it easier to pull dividends received out of a long only account than mark to market gains that feel less real out of my IB account.

Rob
 
globalarbtrader said:
OK to be clear it's currently 40% and used to be significantly less. I've covered some of these reasons before, but:

  • Having 80% of my wealth in a leveraged futures account with a single broker controlled by software written by a complete amateur developer feels risky. It's taken me 10 years to trust this combo enough to move up to 40%.
  • A good proportion of my wealth is in tax shelters, which in the UK at least are difficult if not impossible to use to trade futures. Although some of this can be offset by the higher risk target in my futures account, and therefore better tax efficiency, I would struggle to get near 60% even if I put all my non sheltered investment money into the account. Since I fill my tax shelters to the maximum every year, this constraint will bind more as time goes on.
  • Although I'd like to get higher than 40%, the speed of this will be limited by how quickly I can liquidate long only investments without incurring capital gains tax. Getting to 40% was the most I could do this tax year.
  • As I alluded to in the original post systematic futures trading is quite tax inefficient compared to long term holding of investments where I can time the sale. However the UK capital gains tax free allowance has been reduced to just £3k as of April, so this benefit is worth much less than it was.
  • My equity/bond holdings are long only, but not passive; I do rebalance them (slowly) using simple simple momentum and value rules; so my risk allocation to eg momentum is higher than you think.
  • More of a problem than before, but I preferred the steady drip of long only dividends as an income stream. However in the last 10 years things have changed, my non investment income is higher (it started at zero), and with more of my savings in tax shelters I was receiving much less in dividends anyway; plus with higher interest rates my IB account pays a fair chunk of interest even if the futures portfolio isn't performing. Nevertheless, I know I will still find it easier to pull dividends received out of a long only account than mark to market gains that feel less real out of my IB account.

Rob
More...

if you didn’t have the tax shelter issue would you scale more to your futures trading (even with long term).

would you return stack your long term holdings and futures trading?
 
newwurldmn said:
if you didn’t have the tax shelter issue would you scale more to your futures trading (even with long term).

would you return stack your long term holdings and futures trading?
More...

The theoretical optimal weight on the futures trading is probably somewhere between 50% and 75%, so yes if I didn't have these other issues I would allocate more.

"Return stack" - I see this phrase a lot, I have no idea what it means.

Rob
 
1) When you increased your futures account, did you pay attention to the low expected risk level of your portfolio? Or low level of leverage?

2) How do your UK shares fit into the 40/5/55 allocation? Is this part of "55"? What part?
 
AlexCh said:
1) When you increased your futures account, did you pay attention to the low expected risk level of your portfolio? Or low level of leverage?

2) How do your UK shares fit into the 40/5/55 allocation? Is this part of "55"? What part?
More...

1) Huh?

2) It's in the 55%, and it's about half of it.

Rob
 
You must log in or register to reply here.
Back
Top