Do you "hedge" your mechanical system?

[indexbandit]

Background: I run mechanical systems on market indices and have a robust system which has provided good returns. I am looking to expand my system to included fixed income and metals products and raise more capital. My financial backers are quite happy with our current returns, but are reluctant to provide more capital. The main reason is because the portfolio is basically unhedged, and the positive correlation of the products I'm trading leave me too "all in". The majority of the trading is done intraday, with holding some small positions for extended periods.

So the question is: do you hedge your day-trading portfolios? How? I've considered buying OTM options, but that insurance is quite expensive relative to the markets I trade and the frequency of moves that would require such insurance. While the attempt to diversy to other products is the first step, because my systems are mechanical, there is no guarantee that I would be in divergent positions should some black swan type event occur.

While I do use stops, they would be rendered useless in a catastrophic event.

Do others have this "problem"? What do you do? If you don't , how would you go about convincing investors that this is a non-factor? Any advice from the money-management crowd would be appreciated.

 

[BerkshireBull]

While I do use stops, they would be rendered useless in a catastrophic event.

Do others have this "problem"? What do you do? If you don't , how would you go about convincing investors that this is a non-factor? Any advice from the money-management crowd would be appreciated.

[the note below is about correlation, I'm not sure what to suggest for the money-management aspect]

Rather than hedging my portfolio, I just check the correlation between each instrument on a fortnightly basis. Say I found two instruments that were both closely positively correlated, I wouldn't trade them both in opposite directions (i.e. long on one, and short on the other).

This does cause a dilema though when trading a mechanical system as sometimes - depending on your strategy - you need to take all the trades whether losing or winning, to make sure that the strategy you trade is as close to your backtest results as possible.

Certainly in most 'black swan' events, everything seems to correlate to '1' anyway!

 

[indexbandit]

Thanks for the response Bull. I guess what I meant to say was that I was looking to see if there were any others managing OPM who have a similar problem, rather than just an individual trader with his own cash (and balls) on the line.

As far as the correlations go, its difficult because sometimes the signals are just too damn good to ignore. For example, I may get what I call the "triple lindy" , which is basically a signal that goes off on ES, ER2, and YM at the same time. I have found this to be a highly probable trade and very profitable. So what to do when I get a triple lindy buy signal-- and then a plane crashes into a building? I realize I am being highly dramatic for effect but the question remains-- how to hedge this situation? I may have no other signals on negatively correlated instruments and thus not be in trades which could potentally naturally hedge the situation. I am not a quant guy but have a decent enough understanding. And I do think this is such a low probability, but how do I convince my institutional investors of this?

 

[grantx]

Index,

If you trade all three following your signal, then reduce the risk by trading just one or two. Obviously, one has the least risk but if you also trade in equal size, then the return wil be only one third.

Difficult one.

Grant.

 

[indexbandit]

Index,

If you trade all three following your signal, then reduce the risk by trading just one or two. Obviously, one has the least risk but if you also trade in equal size, then the return wil be only one third.

Difficult one.

Grant.

Thanks Grant. The problem isn't that I'm trading too big relative to my account size-- in fact I usually can't get enough ER2 or YM or NQ- to satisfy. ES no problem. I use different stop parameters for each instrument and I have no problem whatsoever with risk during "normal" trading days. The problem occurs because of the correlation of the products and the nature of mechanical systems i.e the system is making the trading decisions, not the trader (me). In an abnormal market, i could potentially be wiped out and owe more on the account if some sort of black swan occured. While everyone would take a bath in this situation on equities, others might make money as well say being long oil or gold or bonds etc. to mitigate their losses.

So what I am asking is how mechanical traders hedge that risk. If you have a black box system do you even think about catastrophic events? do I need to join a multi-strategy firm or is there an economical way to hedge? Any big boys have an answer?

 

[grantx]

Index,

If an inappropriate hedge is taken then it could actually neutralise your position, ie regardless of wild swings, your net is zero. A rule of thumb is, the greater the protection, the lower the return.

Options are the obvious answer because they allow control of exposure more precisely. In brief, if your long, sell calls; if your short, sell puts. While not vital, it would be preferable to sell options when premiums are ’rich’ (expensive). This is determined by comparing the current implied volatility to previous levels. Note, I’m referring here to historic implied volatility, not historic volatility derived from high/low prices.

If you are doing as well as you say, then it’s only a matter of time when you won’t need the money of others.

Grant.

 

[indexbandit]

Thanks again Grant. All very valid points. Personally, I don't see this predicament as a problem, but my investors do. Primarily, they are concerned with downside risk. While my system has a short bias, I understand their line of reasoning. Again, I can always buy OTM options, but was just looking for a less expensive way and thought others might be able to share what they do.

As far as OPM goes, I will always use others money when given the opportunity as futures trading can be inherently risky and I prefer to diversify the majority of my own funds into longer tem, stable investments.

 

[r.w.]

Indexbandit,

Your problem is very real to me, as a close friend of mine got wiped out last year doing the same thing as you are doing - managing other people's money using a mechanical index futures system. He is now paying his dues in the salt mines (working for a hedge fund). So don't end up in the same situation, ok?

I may be wrong, but to me it sounds like you need a strategy review. Are you saying that you are taking trades in correlated instruments without checking correlations, and then you're thinking about hedging that with oil or gold? Or possibly adding metals and fixed income because the mechanical rules work so well?

My friend's conclusion, after a long deep think, was to reduce leverage (Grant's first post). If that is the conclusion for you too, it would mean telling your backers that the returns so far have been inflated due to over-leverage, and that they cannot expect the same in the future.

Or maybe your edge justify the returns, and then the question is how to best realise it. Futures, options, correlations, etc (Grant's second post), But trying to include more instruments (metals, fixed income) without doing your homework wouldn't be advisable. Hedging is very strategy dependent, so you would have to dissect your strategy in detail first before getting into hedging and managing risk. And mechanical systems is a treacherous field. Be careful out there.

Sorry to be a killjoy, or for possibly misunderstanding, but better that than having you in the salt mines.

 

[r.w.]

Personally, I don't see this predicament as a problem, but my investors do. Primarily, they are concerned with downside risk.

Are you saying that you don't need hedging, expect for marketing purposes? Then why not contractually relieve them of the risk and forget about hedging?

 

[indexbandit]

Hi R.W., thanks for your comments. I did not come here for pats on the back, I'm looking for real responses, regardless of how painful they be. So don't worry about being a killjoy lol. And what happened to your friend is exactly what I am trying to avoid.

My returns are the result of having a verifiable edge (roughly 80% positive trades with roughly a 3:1 reward/risk) plus the inherent leverage that is built into futures contracts, and not the additional cash leverage. My sharpe is around 8. My investors run a fund of fund type approach, and raise capital for new strategies as they discover them. But they do not manage the fund of funds, they rely on the money managers of the investors to do the allocations. So in other words if they get 100K from a client to invest in my fund, they don't necessarily know if that represents 1, 2, 5, or 10% of that investors individual portfolio. They leave that decision up to the investors money manager.

My approach from the beginning has been that this is a high risk/high reward venture and that in no way should it be a large part of an investment portfolio. As I mentioned to Grant, I only keep 10% of my overall in this fund. So when I say I don't "need" hedging, I am merely suggesting that the diversity of an overall porfolio should be hedge enough. Would you risk 2% of your overall portfolio, to potentially make 10%?

Because of the way I started my fund, I have had a ton of autonomy and a sweetheart deal. I wish to maintain the current operation and terms, just expand the scope i.e raise more capital. As far as the correlations go, they are constantly changing. I am not suggesting that metals or fixed income is negatively correlated to stock indices. I don't change my sytem based on these changes. So hedging is a way that will placate my investors. And not just talking about daily hedging-- i do use stops-- I'm talking about hedging against catastrophe and a total market meltdown. No way in hell I would contractually relieve my investors of risk. That would mean that I assume all of the risk. In that instance, what do I need them for?

My original question was how to hedge automated systems. I could buy OTM options and collar a trading range, but they are surprisingly expensive considering that they go unexercised 98% of the time. And continue to do that quarterly until the next black swan, but my intuition tells me that the amount of money lost on those options is going to be greater than any potential benefit they would provide in the next catastrophic event.

 

[grantx]

Index,

The following touches on my point regarding greater protection, lower return.

Buying options is an expense, ie you pay the premiums. These will reduce your overall risk but increase your profit breakeven point.

Selling options means you receivethe premiums, ie an additional source of income for no outlay (assuming you’re covered, which you are), which increases your overall yield. The downside here is that they cap gains (which I don't consider bad for your situation).

Re RW’s friend who got wiped out. This is my interpretation of why top traders get wiped out, and a solution to avoid this. The following is hypothetical.

Your fund is $100,000 and you trade 10 contracts a time (1 lot per $10,000 of the fund), each representing a value of $10,000 .

Your risk is the value per contract of $10,000 (for simplification I’m assuming total potential risk is 100%; we know it’s almost unlimited).

Assume your good trading doubles your fund to $200,000
The return on the fund is 100%
Your profit/earnings is $100,000.

You now trade 20 lots a time and your risk is still $10,000 per contract. One day, there are co-ordinated terrorist attacks on all international exchanges, central banks and airports (OK, maybe I watch too many films). The markets tank and you’re wiped out to the tune $10,000 (100%) per contract or $200,000.

Now, again assume your fund is $200,000 but rather than trading at 1 lot per $10,000, you trade 1 lot per $20,000; your risk per contract is still only $10,000: 10 lots x $10,000 = $100,000

Once more, you are wiped out 100% on your contracts: 10 lots x $10,000 = a total loss of $100,000. But you still have $100,000 in the pot.

Again, your fund is now $200,000 and you are trading 10 lots.
As with your first $100,000, you make a 100% return on your positions
Once more, your overall gain is $100,000 but this equates to a lower percentage return of 50%.
The conclusion would be performance is slipping. But you’ve still made $100,000 and reduced your risk by 50%.

For me, the important figure is the monetary return ($100,000) not the percentage return. Your investors (and possibly yourself) should look at the fund as if they are investing in fixed-income with a constant yield. They can’t have it both ways.

To increasing the overall return further, convert your funds into US Gov bills/notes/bonds for the income stream (yield). Then write (sell) options against these. This will be your (covered) margin/collateral. The options will provide downside protection. The haircut will be minimal, if non-existent.

Grant.