Strategy Guide Note: This article was written before the advent of Commission Free Saturday. As a result, it has been slightly overtaken by events. Now, you might as well re-short everything each Saturday (provided you have the patience), because there is no penalty for doing so. Of course, if you don’t have access to the internet on Saturdays, the analysis remains valid (-Huy).

I discussed short selling in last week’s column (link at bottom of page), but a few e-mails, a flurry of Ticker Talk discussion on Friday, some moments of reflection, and a few hours in front of a spreadsheet have convinced me that a much more in depth analysis of short selling is needed. Several traders have touched on or two of the topics here in their columns or on Ticker Talk, but a few of the topics are brand new (I think), and I also thought it would be handy if they were in place. So call this a “clip and save” column.

Short Selling: What is it (feel free to skip if you know)

I had thought this topic was well understood by now, but several e-mails after my last column told me otherwise. Short selling is basically a bet that the price of a security will drop. On Wall Street, this is done by borrowing the shares from someone else (and paying interest), and selling them, securing the loan of shares with a proportion of the shares’ current value kept in a “margin account” and spending the remainder of the sale elsewhere in the market. If you bet right, and the price of the shares drop, you buy the shares back at a lower price, give these shares back to the broker you borrowed them from, and pocket the price difference. Or you can keep shorting, and the reduced price frees up some of the collateral you set aside. If the price goes up instead of down, you have to pay a higher price to buy the shares back, or else you can keep shorting and hope for the best, while adding more cash to your collateral. Short selling offers the benefits of being able to make money no matter which direction a security moves.

For instance, If you think STOCKA is going to drop from its current price of $100, you can short the security. If the price then drops $10 per share, you cover your short position by buying the stock back at $90, and pocket the $10 difference, minus commissions and any interest on the short.

Short Selling on HSX: (don’t skip even if you think you know)

On Wall Street, for ever share bought, there is a share sold. On HSX, since there is no direct relationship between trades, and since setting up a margin account to track the amount of collateral you need to secure your loans would be a pain in the ass, HSX has created a different sort of system. You pay the current price of the security, but any losses from that point count as gains and vice versa. HSX’s database basically just switches around the plus/minus signs in the gain/loss column, counts a short as a sell instead of a buy, and counts the covering of a short as a buy instead of a sell. This initially struck me as quite clever, and it is still a satisfactory workaround to the lack of true shorting on HSX. But a post by the Dread Pirate Roberts on Friday, and a furious burst of Ticker Talk posts which followed, have pointed out some “quirks” of the HSX system ( I would like to thank Pool Boy, This is Spinal Tap, and Fielding Mellish for helping to clarify the issues involved).

As I pointed out above, on Wall Street, as the price of the shorted stock changes, the amount required in your margin account changes as well. If you bet correctly, your cash is freed up to be spent on other things. If you bet incorrectly, you need to put more money into your margin account or risk getting in hot water. HSX has left out this key property of real world short selling, and this has some very interesting ramfications on your strategies for short selling on HSX.

Diminishing Returns on Capital

As I have written about ad nauseum, one of the most useful ways to make money on HSX is to calculate the Returns on Investment of the various stocks you are considering purchasing, based on estimates of their adjustment or delist prices, and the amount of time until the adjustment or delist happens. (Warning: this gets technical) This incorporates compounded interest. A stock priced at $100 with a 10% daily ROI is expected to grow as follows. Day 1 $110, Day 2 $121. Day 3 133.1 Day 4 146.41.

(Yes, stocks never grow in this exact pattern. The thing is, this pattern of growth will eventually end up exactly reaching the stock’s predicted delist or adjustment price. Calculating the ROI gives you a number which can be easily compared to other stocks, and even though we know that it will not predict growth exactly, I challenge you to come up with a more accurate predictive equation, particularly one which allows for easy comparison between stocks)

Notice that each day’s profits have a cumulative effect on the next day’s profits. Your profits are working for you. Every dollar you maket is automatically reinvested in the same stock, allowing profits to build faster. At the end of 4 days, you are expected to have $146.41 worth of capital gaining 10% interest instead of your initial $100.

Compounding also has an impact when a stock price goes down, but the effect is in the *opposite* direction, causing the nominal amount of the profits to *decrease* rather than increase. For instance, assume a stock initially priced at $100 with a 10% negative ROI. Day 1 $90, Day 2 $81, Day 3 72.9, Day 4 65.61. Again, if on the fourth day you have $65.61 invested short, you will gain 10% profits the next day, assuming the growth is constant.

But here is where the peculiarities of HSX shorting kick in. After the 4th day, you are *not* using $65.61 worth of capital. You still have your initial $100 per share investment, plus your $34.31 in unrealized profits, capitalized in the stock, but only $65.61 of it is earning the 10% interest. The unrealized profits, and an equal amount superfluous security, is just sitting there dead, not working for you. Let’s see how that affects your true rate of return:

Notice how the true ROI declines sharply because of the dead capital. Oddly enough, after years of lecturing traders to ignore the price that they paid for their stocks, the advent of short selling makes the price paid extremely important, since it determines the amount of your dead capital.

What can you do about this? Well, the obviously solution is something that I have taken to calling “re-shorting”. You can cover your short position, realizing all unrealized gains and freeing up all dead capital, and then if you believe the short position is still a good one, you can short the stock again. The problem with this is that re-shorting causes you to take a commission hit. Re-short too often, and commissions will eat you alive. Re-short too infrequently, and your dead capital becomes an albatross around your neck, preventing you from using it productively elsewhere.

What is the right balance to strike? The math for calculating this was beyond me, so I solved the problem using a spreadsheet simulation instead. I simulated the re-shorting of securities at different interest rates to determine the optimal time to re-short. I will publish a table sometime in the next few days reporting my results, but it will only be useful for the most profit hungry trader. Those who don’t want to consult a table every time they want to make a re-shorting decision should follow my “20% Rule”. “If your unrealized profits equal more than 20% of your initial investment, re-short”. The true range varied between 10-30% for the different interest rates I tested, with the value usually coming within a point or two of 20% for most ROIs common on HSX (1-3%). Following this rule will usually get you as close to the profit maximizing point as is humanly possible.

Of course, don’t follow the rule to absurdity. If you have made 20% on a security, and there is only 2% more to gain before it delists at a price slightly less than current, a re-short will cost you all of your gains. And if your security has only made 15% profit, and you know it will adjust splendidly downward in an hour, don’t miss other profit potential by waiting for another five percentage point gain (the upcoming bond adjustments strike me as perfect examples of this. I will probably re-short on all of them an hour before they adjust, and use the profits to max out on any bonds I wasn’t able to afford).

The Upside of HSX Shorting

Astute traders will realize that mathematical balance has to kick in somehow. If this is the case when a short is profitable, what happens when a short goes awry, and you lose money? Well, under most circumstances, you can basically use the disparity between your capitalization on your debt to HSX as an interest free loan. For instance:

You make a gamble on STOCKX, shorting it at $10. It adjusts to $25. You are out $15 a share and when you cover your short position, you will have to pay $15 out of your cash for each share you have. Of course, you will get your $10 initial capitalization back, but your account is still $5 per share poorer after you cover your short than before. So why cover? Unless you think holding the short is going to *lose* you more money than you can make using that $5 per share to play the market, there is no reason to cover the short until HSX forces you to at delist. I need to think about this one a bit more to determine exactly when you should refuse to cover your shorts unless forced (for small losses, getting your initial capital back to use productively is better than avoiding the hit from the short), but if a short goes wrong, you might want to ask yourself whether you should bother covering it.

Further Issues

One of the ramifications of the shorting issues discussed here is that even with re-shorting, the commission costs are going to make short selling less profitable than holding a long position in a stock with an ostensibly identical ROI. For instance if shorting STOCKA mathematically offers a 2% ROI, as does holding a long position in STOCKB, you should take the long position, as you don’t need to pay the commission costs of re-shorting to use your profits. Somewhere along the line, I plan on calculating exactly what the amount of the commission hit is, so that you can know the point at which a short is more profitable than holding long ( I know without doing the math that shorting GCLOO, WRYDE, KBASI, and NNOLT, are more than profitable enough).

Ramifications for Rate of Return

The formula for Rate of Return calculations that I have discussed in my columns and in my spreadsheet is misleading when applied to short sells. The only way you will truly approach that rate of return on a short sell is if you are able to instantly reinvest your short profits and your freed up capital, commission-free, in an investment offering an identical rate of return, after every tiny price change in the shorted stock. Needless to say, this is impossible on HSX. Here is an example of what I mean. Take one of the upcoming Oscar options that stand an excellent chance of delisting at zero, say the Best Picture nomination for Elizabeth. It will be priced at $10, and I predict it will delist at $0 on March 28. Under the old rate of return calc, we have a daily ROI of 100% because the formula is messed up at zero. Let’s imagine it delists at a penny so the formula works (since you can’t re-capitalize any profits less than a penny anyway, this isn’t unreasonable). We get a daily ROI of -16.23%. This is insane. The only way you could approach this rate of return would be if every time the price of the option dropped a penny, you re-invested your penny of profits, plus the penny of freed up capital, commission-free, back into a stock offering an identical investment, such as the same stock. If the option moves in larger increments (and it will), you won’t be able to compound your interest as often, and will end up making much less money. Plus, in the real world, the commissions (once HSX starts actually charging them for shorts and covers) will eat you alive.

So if we want to find out what are rate of return is, in order to compare between long buys and short sells, what can we do? There is no perfect answer. An alternative to the formula discussed above is to modify it slightly.

Our old formula was: ((EDP/Current price)^ (1/(# of days until delist or adjustment)))-1

Our new formula is: (1+(profit/basis) ^ (1/# days))-1

“profit” is the amount of money you expect to make. For a short sell, it is |Current price -delist or adjust price| (the bars (||) denote absolute value – the positive distance away from zero – basically just turn minus signs into pluses)

“Basis” is the amount you are capitalized in the stock. For long buys, this is just current price X number of shares. For short sells, If you are thinking about shorting the stock, it is the amount you *will* pay times the number of shares. If you already own the stock, it is the amount you paid plus any profits you have made.

Basically, this modified formula gives you identical results as the previous formula if you are buying long, but gives you a different number when selling short. Let’s look at that Oscar option again as an example.

(1+(10/10)^(1/40))-1 = 1.75%

This is what your daily ROI will be if you short the option at $10, and hold it constantly, without re-shorting, until it delists. So for those of you who were completed flummoxed about the re-shorting discussion above, which caused your brains to explode, this is where your freedom lies. You can use this formula and not worry about re-shorting.

However, if you follow my “20% rule” and re-short everytime your profits equal 20% of your initial investment, you will substantially exceed this 1.75%. In my spreadsheet simulation, under ideal HSX conditions, I would be able to make about 140% more by reshorting at every 20% profits than I would by holding for the duration. Now I don’t expect these conditions to exist, but I would expect to make more than 1.75% per day. How much more? Beats the hell out of me. It depends on the price trend from now until delist. If the damn thing sits at $10 until the Oscars, you will only make the same amount of profits as someone who never considered re-shorting, since neither of you re-shorted. If it drops $3 tomorrow, then stays there for several weeks, then drops again just before the Oscars, you will make some decent money on 2-3 re-shorts. If it drops steadily from now until delist, you will make around the +140% profits I mentioned above.

So my recommendation to those of you who want to play the re-shorting game, is to download my modified spreadsheet, which now contains both formulas. You can use the conservative one as a lower limit and the idealized one as an upper limit. The more the two disagree, the more your should lean toward the conservative estimate. Generally, the greater the ROI, and the farther off the delist or adjustment, the more the two will disagree. There is no way around using your judgment. Good luck.

One thing that can be said about V2, is that it should keep us number crunchers/analysts busy for months.

Short People got a reason to live

Tom