Repeating my question from late in the previous thread:
It seems to me that if you buy a stock, you could come out arbitrarily well-off, but your losses are limited to the amount you put in. But if you short, your payoffs are limited to the current price, and your losses could be arbitrarily big, until you run out of money.
Is this accurate? If so, it feels like an important asymmetry that I haven’t absorbed from the “stock markets 101” type things that I’ve occasionally read. What effects does it have on markets, if any? (Running my mouth off, I’d speculate that it makes people less inclined to bet on a bubble popping, which in turn would prolong bubbles.) Are there symmetrical ways to bet a stock will rise/fall?
It gets very interesting if there actually are no stocks to buy back in the market. For details on how it gets interesting google “short squeeze”.
Other than that exceptional situation it’s not that asymmetrical:
-Typically you have to post some collateral for shorting and there will be a well-understood maximum loss before your broker buys back the stock and seizes your collateral to cover that loss. So short (haha) of a short squeeze there actually is a maximum loss in short selling.
-You can take similar risks on the long side by buying stocks on credit (“on margin” in financial slang) with collateral, which the bank will use to close your position if the stock drops too far. So basically long risks also can be made as big as your borrowing ability.
if you buy a stock, you could come out arbitrarily well-off, but your losses are limited to the amount you put in. But if you short, your payoffs are limited to the current price, and your losses could be arbitrarily big, until you run out of money.
This is accurate.
it feels like an important asymmetry
This asymmetry comes from the fact that prices are non-negative numbers and do not dip below zero.
Effect on the market? Off the top of my head, here is a couple: long-term shorts are more risky than they seem; and shorting penny stocks (stocks with a low price, typically below $5) is also “extra” risky because your upside is small, but your downside is not.
The fact that shorting penny stocks is dangerous isn’t because of their price per se, it’s because they are typically much smaller companies than normal stocks. That means their profits and prices are much more unstable, so are much more likely to double in value in a short period of time than, say, British Petroleum. Also, because they are so small, much smaller amounts of money can change the price of the stock, which makes them more prone to market manipulation or investor exuberance (this is a non-linear effect, some stocks are so thinly traded that just a couple of trades might happen per month, and whichever is last sets the price for it). Even if a penny stock did a reverse split that made their shares $100 each, they would still be just as risky for these reasons. Also, because penny stocks are thinly traded, stops are much less effective for short sellers.
You are talking about why penny stocks are volatile and yes, that’s all true. However it’s also true that the upside/downside asymmetry is especially pronounced for them.
I’m confused. Do you believe that if you took a penny stock and divided its share count by 100, thus multiplying its price by 100, it would be less risky for short sellers simply because of the price? Let’s assume for the sake of argument that it’s current price is in the $3 range, so the fact that the minimum quote is in pennies isn’t a large effect.
Hm, I think you’re right. The high(er) risk for shorts is a function of volatility (or, more generally, distribution shape) and not of the price level.
The price level has its consequences but these tend to be beneficial for shorts.
You usually avoid unlimited liability by placing a stop order to cover your position as soon as the price goes sufficiently high. Or for instance you can bound your losses by including a term in the contract which says that instead of giving back the stock you borrowed and sold, you can pay a certain price.
Note that for volatile assets (the very ones where you feel uncomfortable about unbounded risk), stop orders are not guaranteed to help. Remember, prices are not continuous—there is a discrete sequence of bids. Price can go from below your stop to MASSIVELY above it before your stop order can be executed. Most often this happens on news when a market is closed, but it can occur intraday as well.
The stop order feels hackish, to me. I was thinking along the lines of short squeezes even before I learned their name. But also, if I’m expecting a bubble to burst, I won’t necessarily be surprised if the price rises massively before it does. I’d be looking for limited exposure without having to chicken out.
The contract term sounds like the sort of thing I was looking for.
Accurate, but not asymmetrical. It’s perfectly symmetrical: purchase of an asset for resale has a loss floor and no gain ceiling, sale of an asset (including short sales) has a gain floor and no loss ceiling. For actual transactions in either direction, there is a practical maximum gain/loss, even when there’s not a theoretical one: if a value goes too far out of modeled range, one of the parties will abrogate when not able to pay the ludicrous amount.
For smaller investors making short-term trades (which is illegal if one has inside info, and unwise if not), generally Call or Put options are used. The constraints of payout/loss can get quite complicated fairly quickly by mixing different strike and maturity options.
That’s not actually wrong, but I think it’s highly misleading.
The failure mode in the long case that corresponds to “stock price suddenly skyrockets” in the short case is that the value of whatever currency you bought the stock with suddenly skyrockets relative to other assets. This (1) is extremely rare, corresponding to a very large negative inflation rate, and (2) is generally something you would be happy about overall because you surely have a lot more dollars (or whatever) than you are spending on the stock.
On the other hand, it’s not nearly so unusual for the price of a stock to increase abruptly, and if you’re shorting it you probably don’t have a lot more of it to be happy about the increasing value of. (If you did, you’d just be selling rather than shorting.)
The failure mode in the long case that corresponds to “stock price suddenly skyrockets” in the short case is that the value of whatever currency you bought the stock with suddenly skyrockets relative to other assets.
Um, does this ever happen? Ever? It looks like an imaginary situation.
Besides, your description implies that you don’t want to measure your wealth in money. What do you want to measure it in?
Don’t privilege any given currency—you don’t buy or sell things, you trade commodities. Sometimes that commodity is a currency, sometimes it’s a stock, sometimes it’s an actual thing.
For the trade that is currency exchange, one currency’s hyperinflation is the other’s hyperdeflation. If you traded away USD for Zw$, your loss (as measured by the amount of Zw$ you could have had later for that USD) was near-infinite.
(note: I honestly believe this, but I am presenting it more forcefully than I believe for socratic and exploration reasons).
Interesting. What other approach makes sense? When you stop treating currency as special, all costs are opportunity costs. The only actual loss you experience from spending now is that you can’t spend it later.
your loss (as measured by the amount of Zw$ you could have had later for that USD) was near-infinite
but if everything is just a tradeable good, why do you choose to measure your loss in Z$? Your loss in McDonald’s hamburgers is zero, your loss in some now-out-of-fashion accessory is actually a gain, etc. etc. If you don’t have money, you have no baseline but just a huge matrix of barter ratios. Whether you have a gain or loss (and its magnitude) solely depends on which pair you pick and there is no pair that’s privileged, is there?
Speaking more generally, not all costs are opportunity costs, some are just actual losses. If you want to think of spending your resources (=money=commodities) in terms of consumption and investment then sure, any consumption incurs opportunity costs because it’s not investment and investment can be seen as risky delayed consumption. But that’s just Econ 101 and it works perfectly well with money as well.
Within the investment world yes, cash is just another asset. But you still need a baseline way to measure things and measuring investment returns in bananas or Swiss watches is kinda inconvenient and an excellent way to screw yourself up. What’s the point?
I see. I think you’re treating your varied anticipated future consumption as your “base currency”, which adds a fair bit of complexity over the simpler two-commodity model. (but it matches common intuitions better, I’ll admit).
I’ve never heard of its doing so. That was approximately half of my point (#1 in the above). If you think I was suggesting it’s a thing anyone should be worrying about, then I respectfully advise you to read what I wrote again. If you merely think I should have been more forceful about how unlikely such an event is, you may be right.
that you don’t want to measure your wealth in money. What do you want to measure it in?
Ability to procure things I value. If my bank account stays exactly as it is and prices of food and books and computers and other things I spend money on halves, then the portion of my wealth embodied in my bank account has effectively doubled. If the prices of those other things double instead, then the portion of my wealth embodied in my bank account has effectively halved.
Of course in practice different things’ prices change in different ways. And in practice the relationship between money and those other things I care about stays pretty stable, which is one reason why Thomas’s analysis is highly misleading. And in practice I care about future prices at least as much as about present prices (but present prices are pretty much our best estimates of future prices, at least for well traded assets). So measuring wealth in money works very well in principle. But Thomas was (in effect) envisaging a weird situation in which the value of money relative to everything else increases abruptly, and although it’s very unlikely ever to happen it seemed worth pointing out some actual likely consequences.
[EDITED to add: This is currently at −1. I honestly have no idea why that might be. Anyone—preferably whoever actually downvoted me—want to explain?]
Um, does this ever happen? Ever? It looks like an imaginary situation.
The closest direct analog is a crash—if I go from being able to buy one share for one dollar to being able to sell my one share for one penny, one can see this as the value of cash going up 100X.
(This is somewhat contrived when dealing with cash, but it does seem that the foundational level of wealth is food and ammunition. It could happen that the exchange rate between those and cash and stocks skyrockets, and that would be Bad News for a lot of reasons.)
Indirect analogs rely on opportunity cost—because you invested in A and got a 2X return, you missed out on investing in B, where you would have gotten a 2000X return. This is a profoundly unhealthy way to view markets.
The closest direct analog is a crash—if I go from being able to buy one share for one dollar to being able to sell my one share for one penny, one can see this as the value of cash going up 100X.
For this to work you need for basically all financial assets to crash, not just some particular stocks. Besides, we still have the problem of the unit of measurement. If you want to measure your wealth in consumables (say, cans of beans) then for “unlimited” losses from long positions you need not only a financial crash, but also cans of beans becoming really really cheap. This is.. unlikely.
All in all, there is a real asymmetry between going long and shorting. Trying to construct imaginary situations in which you could lose a lot from being long isn’t terribly helpful.
because you invested in A and got a 2X return, you missed out on investing in B, where you would have gotten a 2000X return. This is a profoundly unhealthy way to view markets.
I think it is the correct way to view the markets once you add risk management. If the probabilities of getting those returns for A and B were the same (and the distributions were shaped the same), you indeed missed out greatly.
Yeah, basically the only scenario I see is cans of beans becoming very cheap in terms of ammunition for unethical reasons.
I think it is the correct way to view the markets once you add risk management. If the probabilities of getting those returns for A and B were the same (and the distributions were shaped the same), you indeed missed out greatly.
Agreed—I’m making the assumption that such comparisons are made retrospectively instead of prospectively, and thus are implicitly ignoring risk.
the only scenario I see is cans of beans becoming very cheap in terms of ammunition for unethical reasons.
Unethical even in the Zombie Apocalypse scenario? X-)
But sure, if the entire financial system {im|ex}plodes, your shorts aren’t going to do you any good and so we finally achieve symmetry—everyone is fucked.
I’m making the assumption that such comparisons are made retrospectively instead of prospectively
It is still the right way even retrospectively if you think in probability distributions. And, of course, anything “ignoring risk” is automatically the wrong way to think about the markets :-)
So I can trade one currency for another, and then trade back, and the amount I now have in the first currency can be arbitrarily high. This doesn’t feel like it particularly changes anything.
Repeating my question from late in the previous thread:
It seems to me that if you buy a stock, you could come out arbitrarily well-off, but your losses are limited to the amount you put in. But if you short, your payoffs are limited to the current price, and your losses could be arbitrarily big, until you run out of money.
Is this accurate? If so, it feels like an important asymmetry that I haven’t absorbed from the “stock markets 101” type things that I’ve occasionally read. What effects does it have on markets, if any? (Running my mouth off, I’d speculate that it makes people less inclined to bet on a bubble popping, which in turn would prolong bubbles.) Are there symmetrical ways to bet a stock will rise/fall?
It gets very interesting if there actually are no stocks to buy back in the market. For details on how it gets interesting google “short squeeze”.
Other than that exceptional situation it’s not that asymmetrical:
-Typically you have to post some collateral for shorting and there will be a well-understood maximum loss before your broker buys back the stock and seizes your collateral to cover that loss. So short (haha) of a short squeeze there actually is a maximum loss in short selling.
-You can take similar risks on the long side by buying stocks on credit (“on margin” in financial slang) with collateral, which the bank will use to close your position if the stock drops too far. So basically long risks also can be made as big as your borrowing ability.
This is accurate.
This asymmetry comes from the fact that prices are non-negative numbers and do not dip below zero.
Effect on the market? Off the top of my head, here is a couple: long-term shorts are more risky than they seem; and shorting penny stocks (stocks with a low price, typically below $5) is also “extra” risky because your upside is small, but your downside is not.
The fact that shorting penny stocks is dangerous isn’t because of their price per se, it’s because they are typically much smaller companies than normal stocks. That means their profits and prices are much more unstable, so are much more likely to double in value in a short period of time than, say, British Petroleum. Also, because they are so small, much smaller amounts of money can change the price of the stock, which makes them more prone to market manipulation or investor exuberance (this is a non-linear effect, some stocks are so thinly traded that just a couple of trades might happen per month, and whichever is last sets the price for it). Even if a penny stock did a reverse split that made their shares $100 each, they would still be just as risky for these reasons. Also, because penny stocks are thinly traded, stops are much less effective for short sellers.
You are talking about why penny stocks are volatile and yes, that’s all true. However it’s also true that the upside/downside asymmetry is especially pronounced for them.
I’m confused. Do you believe that if you took a penny stock and divided its share count by 100, thus multiplying its price by 100, it would be less risky for short sellers simply because of the price? Let’s assume for the sake of argument that it’s current price is in the $3 range, so the fact that the minimum quote is in pennies isn’t a large effect.
Hm, I think you’re right. The high(er) risk for shorts is a function of volatility (or, more generally, distribution shape) and not of the price level.
The price level has its consequences but these tend to be beneficial for shorts.
You usually avoid unlimited liability by placing a stop order to cover your position as soon as the price goes sufficiently high. Or for instance you can bound your losses by including a term in the contract which says that instead of giving back the stock you borrowed and sold, you can pay a certain price.
Note that for volatile assets (the very ones where you feel uncomfortable about unbounded risk), stop orders are not guaranteed to help. Remember, prices are not continuous—there is a discrete sequence of bids. Price can go from below your stop to MASSIVELY above it before your stop order can be executed. Most often this happens on news when a market is closed, but it can occur intraday as well.
The stop order feels hackish, to me. I was thinking along the lines of short squeezes even before I learned their name. But also, if I’m expecting a bubble to burst, I won’t necessarily be surprised if the price rises massively before it does. I’d be looking for limited exposure without having to chicken out.
The contract term sounds like the sort of thing I was looking for.
You can always play with options to construct whatever payoff structure you desire.
Accurate, but not asymmetrical. It’s perfectly symmetrical: purchase of an asset for resale has a loss floor and no gain ceiling, sale of an asset (including short sales) has a gain floor and no loss ceiling. For actual transactions in either direction, there is a practical maximum gain/loss, even when there’s not a theoretical one: if a value goes too far out of modeled range, one of the parties will abrogate when not able to pay the ludicrous amount.
For smaller investors making short-term trades (which is illegal if one has inside info, and unwise if not), generally Call or Put options are used. The constraints of payout/loss can get quite complicated fairly quickly by mixing different strike and maturity options.
You never only buy, but at the same time you have traded your dollars, euros or whatever currency for that stock.
There is nothing like “buying” and “shorting”—it’s always trading. Swapping two “currencies”.
That’s not actually wrong, but I think it’s highly misleading.
The failure mode in the long case that corresponds to “stock price suddenly skyrockets” in the short case is that the value of whatever currency you bought the stock with suddenly skyrockets relative to other assets. This (1) is extremely rare, corresponding to a very large negative inflation rate, and (2) is generally something you would be happy about overall because you surely have a lot more dollars (or whatever) than you are spending on the stock.
On the other hand, it’s not nearly so unusual for the price of a stock to increase abruptly, and if you’re shorting it you probably don’t have a lot more of it to be happy about the increasing value of. (If you did, you’d just be selling rather than shorting.)
Um, does this ever happen? Ever? It looks like an imaginary situation.
Besides, your description implies that you don’t want to measure your wealth in money. What do you want to measure it in?
Value of USD (and many other currencies) against Zimbabwe Dollar skyrocketed pretty spectacularly in July 2008. I have a Z$10^12 note at home.
This is a rare, but absolutely possible, outcome.
I don’t see how it fits the case.
If your domestic currency is USD, you bought an asset (foreign currency) and that asset dropped in price to near zero.
If your domestic currency is Zimbabwe dollar ZB, then you had to *short the USD to suffer huge losses.
As gjm notes, you need a huge negative inflation rate (aka deflation) and I don’t think that ever happened, at least during the fiat money era.
Don’t privilege any given currency—you don’t buy or sell things, you trade commodities. Sometimes that commodity is a currency, sometimes it’s a stock, sometimes it’s an actual thing.
For the trade that is currency exchange, one currency’s hyperinflation is the other’s hyperdeflation. If you traded away USD for Zw$, your loss (as measured by the amount of Zw$ you could have had later for that USD) was near-infinite.
You are now talking, basically, opportunity costs. I don’t think your approach makes sense.
(note: I honestly believe this, but I am presenting it more forcefully than I believe for socratic and exploration reasons).
Interesting. What other approach makes sense? When you stop treating currency as special, all costs are opportunity costs. The only actual loss you experience from spending now is that you can’t spend it later.
Well, to start with the Z$ example, you say
but if everything is just a tradeable good, why do you choose to measure your loss in Z$? Your loss in McDonald’s hamburgers is zero, your loss in some now-out-of-fashion accessory is actually a gain, etc. etc. If you don’t have money, you have no baseline but just a huge matrix of barter ratios. Whether you have a gain or loss (and its magnitude) solely depends on which pair you pick and there is no pair that’s privileged, is there?
Speaking more generally, not all costs are opportunity costs, some are just actual losses. If you want to think of spending your resources (=money=commodities) in terms of consumption and investment then sure, any consumption incurs opportunity costs because it’s not investment and investment can be seen as risky delayed consumption. But that’s just Econ 101 and it works perfectly well with money as well.
Within the investment world yes, cash is just another asset. But you still need a baseline way to measure things and measuring investment returns in bananas or Swiss watches is kinda inconvenient and an excellent way to screw yourself up. What’s the point?
I see. I think you’re treating your varied anticipated future consumption as your “base currency”, which adds a fair bit of complexity over the simpler two-commodity model. (but it matches common intuitions better, I’ll admit).
I’ve never heard of its doing so. That was approximately half of my point (#1 in the above). If you think I was suggesting it’s a thing anyone should be worrying about, then I respectfully advise you to read what I wrote again. If you merely think I should have been more forceful about how unlikely such an event is, you may be right.
Ability to procure things I value. If my bank account stays exactly as it is and prices of food and books and computers and other things I spend money on halves, then the portion of my wealth embodied in my bank account has effectively doubled. If the prices of those other things double instead, then the portion of my wealth embodied in my bank account has effectively halved.
Of course in practice different things’ prices change in different ways. And in practice the relationship between money and those other things I care about stays pretty stable, which is one reason why Thomas’s analysis is highly misleading. And in practice I care about future prices at least as much as about present prices (but present prices are pretty much our best estimates of future prices, at least for well traded assets). So measuring wealth in money works very well in principle. But Thomas was (in effect) envisaging a weird situation in which the value of money relative to everything else increases abruptly, and although it’s very unlikely ever to happen it seemed worth pointing out some actual likely consequences.
[EDITED to add: This is currently at −1. I honestly have no idea why that might be. Anyone—preferably whoever actually downvoted me—want to explain?]
The closest direct analog is a crash—if I go from being able to buy one share for one dollar to being able to sell my one share for one penny, one can see this as the value of cash going up 100X.
(This is somewhat contrived when dealing with cash, but it does seem that the foundational level of wealth is food and ammunition. It could happen that the exchange rate between those and cash and stocks skyrockets, and that would be Bad News for a lot of reasons.)
Indirect analogs rely on opportunity cost—because you invested in A and got a 2X return, you missed out on investing in B, where you would have gotten a 2000X return. This is a profoundly unhealthy way to view markets.
For this to work you need for basically all financial assets to crash, not just some particular stocks. Besides, we still have the problem of the unit of measurement. If you want to measure your wealth in consumables (say, cans of beans) then for “unlimited” losses from long positions you need not only a financial crash, but also cans of beans becoming really really cheap. This is.. unlikely.
All in all, there is a real asymmetry between going long and shorting. Trying to construct imaginary situations in which you could lose a lot from being long isn’t terribly helpful.
I think it is the correct way to view the markets once you add risk management. If the probabilities of getting those returns for A and B were the same (and the distributions were shaped the same), you indeed missed out greatly.
Yeah, basically the only scenario I see is cans of beans becoming very cheap in terms of ammunition for unethical reasons.
Agreed—I’m making the assumption that such comparisons are made retrospectively instead of prospectively, and thus are implicitly ignoring risk.
Unethical even in the Zombie Apocalypse scenario? X-)
But sure, if the entire financial system {im|ex}plodes, your shorts aren’t going to do you any good and so we finally achieve symmetry—everyone is fucked.
It is still the right way even retrospectively if you think in probability distributions. And, of course, anything “ignoring risk” is automatically the wrong way to think about the markets :-)
So I can trade one currency for another, and then trade back, and the amount I now have in the first currency can be arbitrarily high. This doesn’t feel like it particularly changes anything.
You are welcome!