Search results

I do the two normal things: Forecasts do their thing, and obviously reduce a position if the forecast is weakening. If I loose money, I recalculate the portfolio for the new, reduced portfolio (which basically means getting rid of some contracts). (Kelly). These two things are the basic stuff...

Don't jinx it! :) I've already given almost half of it back in the last couple of days (7.2% DD from HWM @ 25% vol).

Well I'm trying to build an HFT market making algo based on the paper here https://arxiv.org/pdf/1105.3115.pdf (using the results at the bottom of page 13). The issue here is that it's tough to paramaterise the equation. I can do a reinforcement type learning and get it to learn from trading...

How? I think the issue might be related to this: https://en.wikipedia.org/wiki/Multi-armed_bandit

Yes exactly, but where you can only use live trading to inform you of the parameters. I'm conscious of using a naive method that may overfit. I can handcraft a prior, but still need a reasonable posterior.

@globalarbtrader is there a preferred method for learning parameters in an online fashion, where a backtest isn’t possible?

Isn’t that vol is the std deviation of returns? So for example, I expect a Sharpe of 0.8 on vol of 25%. So my expected 1std profit and loss for the year is 20% +/- 25%, not just 25%. 2std would leave me in the range -30%..+45%, a 1 in 30 year occurrence (theoretically). Secondly, I do my back...

These stocks are more likely to have a drawdown when risk normalised, aren't they?

Is that a mathematical thing or a psychological thing? Would a regular coupon from your trading capital hamper you in the long term?

I've thought long and hard about this problem. Here's what I've tried: * @globalarbtrader 's method - bootstrap portfolio weights to optimise overall Sharpe ratio. Possible risk: overfitting (e.g. heavy equities, heavy VIX) * Some kind of max separation based on the correlation matrix of weekly...

Hahaha I hope so! But I really think the backtests are flawed, or otherwise, a lot of the overall uptrend of crypto is leaking into the results. On an aside note though, for anyone holding Palladium futures, it's been a great week!

Actually, the backtests are currently coming out with a Sharpe of 8, so I'm inclined to think there's something wrong with them. It's quite challenging to backtest market making though. I've got a naive simulation (similar to what I did for futures/trend following). I'll try to build a more...

I think once you approach a Sharpe of ~1, there isn't much you can do to really improve things on a daily strategy (correct me if I'm wrong!), even if you wrote a whole extra strategy and diversified into it (e.g. AHL Dimension). For the most part, I'm just leaving it alone. At some point I'd...

I think this chart could be easily misinterpreted. The performance is very strong for the first five years. Rebase the graph at 1990, what would you see? Secondly, these lines don't look they've had the volatility normalised, so I'm not sure it makes sense to compare them on a chart like this...

It's not statistically significant. I also added the rule, and saw that there wasn't the statistical evidence to include it, so removed it. I did implement @globalarbtrader 's short bias on vol futures.

Seems sensible, good choice of instruments. For eurodollar, can I check you're trading many contracts away (I trade 900 days out)? and for v2tx, you're not in the front contract?

Definitely. What instruments are you trading?

Just those two. I got around 0.7 as well, with equal weights in the portfolio. Then I bootstrapped the portfolio to find better weights, and that improved things. I'm a little concerned about overfit (I don't like this method), but I don't think equal weights is right either. You can see my...

I did some work on this. Two of the obvious options are: Mean reversion over shorter time scales. Trends tend to work on periods of more than a month, mean reversion in the shorter term. Trading a spread - e.g. long nasdaq short s&p. I couldn't get either to work really: the short time scale...

@globalarbtrader what are your thoughts on doing 'sell and hold' on VIX/VSTOXX, instead of EWMAC? This produces significantly higher Sharpe on backtest, and also seems to make 'fundamental sense' - or am I missing something?