Tuesday, October 1, 2013

Dynamic Hedging

By definition, hedges are supposed to lose money to allow you to have more aggressive long exposure. Your extra exposure is supposed to gain more than your hedge loses but in the even tthat you are wrong, the extra exposure will roughly cancel out, or even decline less than your bearish bet.

I believe however that you can do better. There are bearish market bets in an uptrend that I also call hedges for lack of a better term, that actually can make money regardless of what market does, while your bullish plays make money. The trick is knowing specific setups that are mathematically expected to decline based upon history, even in a bull market, but even more so in a bear market. Once you know these setups, with excellent execution, you can make money from a bearish bet while the market works to the upside, or make even more from that setup as it declines. I prefer the setups that are very common as opposed to the ones that have worked the best with a smaller sample size. This way I can always find a trade and be more confident that it will work, and also test more recent results in a shorter amount of time. If I only need one set up and can find 3, I can use other criteria to try to get a better return. I prefer the swing trade type of setups such as those from a candlestick chart as opposed to looking for price patterns.

If all you did was have an even amount of bullish plays and bearish trades or all you did was "pairs trading" and pulled this off, you could gain with almost zero market correlated risk, (loss from market direction) and minimal account volatility, which would allow you to heavily and aggressively deploy your cash for maximum gain, or keep the same high levels of cash for maximum stability.

However, the way that I like to hedge is by some ratio of something like 2 bearish bets for every 3 bullish bets (Ratio should be set so that if I deploy all of my cash, I don't exceed my desired maximum % long exposure). The alternative that I like even more is to maintain some sort of allocation percentage on the long side, and use hedges only when the long side trades are working and suddenly have a larger percentage of capital than you intended. For example, if you have 20% short term trades, 20% long term trades, and 10% stock index and 20% low risk income with 30% cash, you are looking for about 50% long exposure. Now say the short term trades explode to the upside and are worth 30% of your portfolio putting you at 60% long exposure, now you would possibly want to offset 10% of it because you are now beyond your intended ratio of 50% risk, 20% stable income and 30% cash. You could sell your 10% stock index, but these are short term trades. Because they are short term you may not opt to hedge at all since you will be hitting your targets and selling soon. You may just sell early, but really I would look at two options (in this order)
1)Waiting until you hit target or selling any stocks already near, at, or past the target.
2)Short term hedging

This second way when I start winning I protect my gains a little better and typically if my plays are positively correlated with the market, I will be adding hedges higher after a move to the upside in the market, where it will be easier to spot failed moves and also get long the moves that have yet to work.
In some cases, the setups will be so good and the timing will be so right that you have to take on new positions even though you may not want additional market risk exposure. In that case, you simply will match one hedge per one bullish bet.

In the case of short term hedging after offsetting your "balance" in the example above, you would aim to maintain around 10% of short term hedges. However, the exit strategy on these hedges are actually NOT going to be just when you take your winning trade or bullish trade off, but are going to have a specific holding period that has proven to work in the past with a possible extension. I might go 3 or 5 days with a possible extension to 10 days.

I do not do "long term hedging" for the most part because I don't find a ton of success with bearish price patterns or have enough experience with them, and candlestick patterns are usually short term trading vehicles. Instead, I will just take off a hedge and then add one that same day. For example, if you go from the initial example of 20% short term trades, 20% long term trades 10% stock index, 20% income 30% cash and end up with 30% long term trades? The decision in order will go like this.
1)Take off short term trades at, past, or near target (or that stop out) and don't put on any new short term trades. (In this example, you would aim to get to 10% short term trades to offset long term, but only if possible without selling stuff early)
2)Reduce or take off stock index investment allocation. Most likely you will be selling high as market would likely have rallied to get long term trades to the inflated level.
3)Sell any long term trades that are near, at, or past their target.
4)Put on short term hedges to offset risk and continue to replenish old hedges that you sell with new that you buy until the first three options are able to offset the extra risk and reduce or eliminate the need for hedging.
5)Put on long term index fund hedge (last resort option)

I do not think you should just stick the cash into an index fund on a bearish bet if you can since in this case you are only betting on your one trade to perform well enough to offset your losses, rather than betting on BOTH to be winning trades regardless of market direction, but even more winning if the market moves in their respective directions (which are opposite of each other). It is probably better to put on an index hedge then to have too much risk, but this will reduce your gains relative to the other strategies.

Now the "dynamic" part of the hedging is that the actual "percentage" exposure to risk that you are aiming for will and should fluctuate according to your overall market outlook but drawing exceptions when individual setups or timing is good enough.

For example if you have no idea on individual market direction, a 50% bullish, 50% cash split is best, and if it goes to say 60% bullish, you put up 10% hedge, if it goes to 70% you put up a 20% hedge.

If market direction is expected to be say at a 60% probability of moving an equal amount to the upside, you want to be 60% bullish, so you might have to take off some hedges or add some longs so that you have that long exposure. Either 60% bullish no hedge, 70% bullish 10% hedge or some combination.

If market direction is expected to be at say a 40% probability of moving an equal amount to the upside, you want to be 40% bullish. (or 50% bullish, 10% hedged). At this point you may even wish to be aggressive and be something like 40% bearish with a 10% bullish hedge, but I would never want to be beyond even 40% bearish under any circumstance because of things like borrowing costs, margin calls, puts being more expensive and typically pricing in the borrowing costs, puts being more expensive in a bear market than they are cheap in a bull market, etc. For that reason I would not flip to the theoretically correct 60% bearish instead, and the aggressive nature of the shift also would cause a lot more fees, turnover and psychologically would be a difficult adjustment for many.

I personally would not go beyond 60% long or under 40% long, and I would not go over 40% short (or the equivalent, or a total of 50% short (even if matched with 10% or 20% long). That is about recognizing your limitations with regards to market direction, and should keep you focus less on market direction and more on individual setups and relative out-performance or achieving "alpha".

Now if you use leverage the strategy should be much different. Your cash position must be much larger. If you have a 8% stop for example in stock, 1% risk would put about 12.5% of capital towards the trade. With the option, you instead might position 1% of your capital towards the trade. Big difference. As a result, you should really be dividing any position by 12.5% to determine your leverage. If you are more aggressive and risk 2% per trade, you are taking 25% of capital allocated towards a trade or 2% towards an option.
I think though, you might be able to be a bit more aggressive than that using leverage since you can create more positions and be more diversified and find more trades at a lower correlation. As a result you might only divide by 10.

In other words, the option equivalent for a strategy of something like  20% income, 30% cash 20% long term, 20% short term, 10% stock index might be 40% income, 55% cash, 2% long term, 2% short term, 1% stock index. Since you have more non risk cash, you can put more into stable income. Since you have more stable income and cash you might increase the leverage further. Afterall, the only reason we go into cash is for rebalancing and protection against being wrong, not because we can't afford more aggressive risk.

If I was to use leverage, I personally would use some of that cash to leverage some no or low correlated plays. As described in some previous strategies, I could go 2% currency, 2% bonds, 2% commodities, 2% stock index. Because of this allocation being basically zero correlation with some inverse correlation names, and because of the ability to reduce stock index when individual position sizes get too large, This is another way you can give yourself maybe HALF of the position size of this low correlation and apply it to each of the short and long term trades. In other words, 36% income, 50% cash, 3% long term, 3% short term, 2% stock index, 2% currency, 2% bonds, 2% commodities.



Also, since you have a positive expectation when it comes to your trades, and aren't simply rebalancing a stock index, I believe this favors a more aggressive strategy for EITHER the stock one or the option. Since the option's edge is magnified, this is especially true if your timing is excellent as well. If you have a plan to add cash and your starting capital is lower relative to the cash you will add (adding 1000 a month to an account that is only 10,000) this can allow you to be much more aggressive as well since after a 20% decline, you won't find recouping your losses as difficult as you would without a deposit. A 11% gain on portfolio would get you from 90% back to 100% rather than a 25% gain. A 11% plus another 10% or 21% would get you from 90% back to 110% or recoup what you lost, rather than a 25% gain. That is just on a monthly basis and after the month you have another 1,000 and it will be even easier.

if your strategy is 36% income, 50% cash, 3% long term, 3% short term, 2% stock index, 2% currency, 2% bonds, 2% commodities, then with that cash you will occasionally rebalance, but also will note that if your short term plus long term plus stock index equals more than 8% you hedge by the amount needed to offset, or reduce your stock index position or short or long term position near, at, or beyond the target For example if short term trades double to 6%, you hedge 3%.

This dynamic allocation is very similar to the one described before but I will reillustrate here.


I want to add one that shows you how to respond if you are using leverage.
(typo edit: top right corner should say cash 50%)

What I actually do involves occasionally using leverage, and occasionally using stock, but the stock position can be nearly 10 times the option. So the actual strategy might change dramatically depending on whether or not I have stock and how many, and whether or not I have a bond trade available at the moment. My strategy is not yet entirely defined, but I do have a philosophy that works almost exactly as described, but the actual allocations fluctuate a bit too much at the moment. The reason I don't have a predefined strategy yet, is because I want to find the position sizing strategy that accomplishes my goals, and then structure everything else around it including: Allocation(commodities/stock index/bonds trades/currency trades), income, cash, and hedging. These all are structured AROUND the long and short term trades. In other words, the allocation strategy needs to be such that the stock index can be large enough to offset the potential growth of the long term trades (stock index size should be anywhere from 2/3 to 3/2 the size). The remaining allocation should typically EQUAL the stock index position size (unless it's taken off to reduce risk), and the cash then must be large enough to accommodate the need of liquidity as well as additional cash for volatility risk, and the remaining which is illiquid safety for reduced volatility and capital inflows and also must be smaller than the actual cash percentage. Substituting options for stock is fine, but I would probably prefer 2 half positions (2 5% stock to every 1% option) as opposed to one full position.

Previously, the philosophy looked at hedging as just any hedge such as an index option. With the introduction of  individual short term hedging in positions that are expected to be profitable, I might just by default set up a 1% hedge and increase the short term position by a half a percent and the long term by half a percent. Any additional hedging might be to reduce the risk. This way I am always hedged, but in a way that expects to yield a profit as well as lets me add additional exposure.
Either way, I think a particular ratio up to a certain amount such as 2 hedges for every 3 trades I add is a good way to keep myself in check from over trading and taking on too much risk. The % allocated to income will also limit the amount you can actually get long and provide some liquidity if you absolutely need it and some longer term stability.




No comments:

Post a Comment