Asset Management

My brother thinks of himself as a farmer, which he is. But I think of him as an asset manager. He has chosen to hold his assets in land, tractors, ploughs; and that's him, driving his tractor, pulling his plough over his land, fixing the tractor, fixing the plough, managing his assets. He earns his income doing the 1001 things he needs to do to get the best return from his assets.

I'm the opposite. I left the farm and got a job. I spend some time managing my assets, but much less than my brother. If I spent more time managing my assets I could get a much better return from my assets. But there's only so many hours in the day, so I face a trade-off between the wages I get from my job and the return I get from my assets. Yet someday I will find it makes more sense to quit my job and become a full-time asset manager, like my brother. People will say I have retired.

Assets don't just manage themselves. It would be nice, for people like me, if my income were simply Y = WL + rK. Then I could simply work at my job for a salary WL, and the value of my assets K would grow at some exogenous (to me) rate r, plus whatever I saved from my salary. But r is not exogenous. Just as my labour income WL depends on my effort and skill in managing my human capital, so my non-labour income rK depends on my effort and skill in managing my non-human capital.

And once you start looking at it like that, the distinction between labour and non-labour income becomes highly suspect. Y = WL + rK makes no sense as a description of my brother's income. Y = F(S,K,L) works much better, where S is his skill, K is his assets, and L is the effort he puts into managing those assets.

Assets don't just manage themselves. We either manage our own assets ourselves, or pay someone else to manage them for us. Actually, it's more of a continuum. At one extreme you have my brother, who spends all his time managing his assets. At the other extreme you have someone who puts all his assets into a savings account at the nearest bank because it requires the least skill and effort. And in between you have people like me. I own shares, which means I delegate the management of my assets to the managers of that company. I own mutual funds, which means a double layer of delegation. And my broker advises me on who to delegate to. My sister rents my land and farms it. And one potentially valuable barn sits empty because I don't have the time and skill to manage it properly right now. It's an option, not a barn. One day I will exercise it, when the strike price of my time falls.

Very few people are like my brother, who works at managing his own assets. Most people work at managing other people's assets, and get other people to work at managing their own assets. It's not just people who work in finance, who think of themselves as managing other people's assets. If you work for a company you don't own, you are working at managing someone else's assets. Even my brother is not a pure case. He borrows money from the bank, which means he is managing someone else's assets. And he hires contractors and mechanics, which means he is delegating the management of his own assets to others.

Most of us delegate the management of most of our assets to other people. But that creates a Principal-Agent problem. We won't manage other people's assets as well as we would manage our own. And much of the effort we expend in managing our assets is monitoring our agents, or choosing the right agents.

If all of us have an income function Y = F(S,K,L), what does that function look like?

If you have no assets, so K=0, it probably looks like this: F(S,0,L) = W(S)L. The wage you earn depends on your skill, and your income is then roughly proportional to your effort or hours worked.

If you do no work, and simply find the easiest possible way of investing your assets, so L=0, it probably looks like this: F(S,K,0) = rK where r is the rate of return on something like a savings account or GIC.

There are probably increasing returns to scale in managing assets. If you double your assets, you need to spend more time but don't need to spend double the time managing them to get the same rate of return. Doubling K holding L constant, or doubling L holding K constant, will mean Y increases but less than double. But doubling both K and L will more than double your income.

Rich people probably face offsetting income and substitution effects on effort. The richer you are the more leisure you want to consume. Leisure is a normal good. But at the same time the greater your assets the greater your marginal return from effort in managing them.

What does all this imply about the distribution of income and wealth over time? I'm not sure. Those born with higher skill will earn higher income. That's standard. But they will also have a higher rate of return on their assets, and so will have a greater incentive to save part of that income. So they will tend to get richer still over time. The distribution of wealth and income will be more skewed than the underlying distribution of skill that creates it. Eventually though, people get old and tired, and so can't devote the effort to managing their assets. Or their skills decay. Or the next generation of heirs regresses towards the population mean level of skills, so their rate of return and rate of accumulation slows. Plus, if your assets keep on doubling you won't and can't keep on forever increasing the effort needed to manage them and get the same rate of return. So the long run distribution of wealth and income may remain stationary.

Just some random thoughts on returning from my fortnight holiday/asset management trip to England. I will get back to macro/money now.

Comments

[Patrick]

Hope you enjoyed your holiday Nick.

I'd add another variable to Y=F(S,K,L) - luck. Especially for a farmer. Someone can be lucky and their Y at time t jumps due to luck, feeds back into K, which feeds forward into higher Y forever. Similarly, a farmer (for example), regardless of his skill, can be wiped out by bad weather or disease.

[Frances Woolley]

Nick: "I own shares, which means I delegate the management of my assets to the managers of that company. I own mutual funds, which means a double layer of delegation. And my broker advises me on who to delegate to."

This is an absolutely huge issue. Would you conclude that, as a result of this triple management layer (your broker, the mutual fund manager, the corporate manager) there is little assurance that those corporate assets you own indirectly are well-managed? Or do you figure mutual fund managers exercise more effective restraint on corporate graft than you could?

"I spend some time managing my assets" What readers of this blog don't know is that you are exceptionally good at fixing things, especially cars. And by exceptionally good I mean better than I am.

What I can't work out is: am I just exceptionally bad at fixing things myself (the laser printer that won't work because the paper keeps jamming, the clock radio with a broken on-off switch, the refrigerator that's leaking water)? Or has asset management technology changed fundamentally?

[Frances Woolley]

http://modeledbehavior.com/2011/05/26/corporate-governance-and-the-plutocracy/

[Patrick]

Frances:

http://www.theglobeandmail.com/news/technology/tech-news/microsoft-board-backs-ballmer-over-einhorn/article2035941/

If Karl Smith, or Einhorn for that matter, thinks that maximizing shareholder value has anything to do with it, I'd suggest they have the wrong model in mind. We're talking about a group of close friends here (the M$ board of directors). Ballmer has been with M$ a long, long time and the board are all his long time brothers-in-arms. They all believe they've made each other billionaires. They probably all go to the same country club, their wives shop and lunch together, their kids probably go (or went) to the same private schools. No way are they going to throw one of their own under a bus for the sake of the masses of the great unwashed M$ shareholders. Blood is thicker than water.

[Chris J]

Nick,
Surely the tipping point is when your assets earn enough that you can live off the asset growth and be better off next year? I think this is just asking if your money works for you or if you work for it? And while farmers are held in very high esteem generally - especially the brother of the article I am sure - if you are going to be cold about it, your brother is a business owner. Maybe a capital-intensive business, but a business owner. This reminds me of the one liner: "When you work for yourself, everyday is a Saturday. When you work for yourself, you work Saturdays."

A person living off investments is not tied to the fortune of just one company. (Unless that company is Berkshire Hathoway or the investor is an idiot.) A business person's livelihood is tied to one company.

I won't debate the validity of your model, but I think you are using it to describe two different situaions.

Cheers,

[RSJ]

"And once you start looking at it like that, the distinction between labour and non-labour income becomes highly suspect. "

So does this mean that you will no longer argue that capital income "is not income" for purposes of income taxes? Do we agree that all income, both capital and wage income, should be taxed equally to create the most efficient tax? :)

[Nick Rowe]

Patrick: you are of course right about luck. But whereas most discussions of the rate of return on assets talk about how risk and liquidity affect the equilibrium rate of return, I wanted to abstract from those and talk about effort and skill.

Frances and Patrick: I don't see how there cannot be a significant Principal-Agent problem. Here's an example, right from this post:

My brother grows malting barley (for beer). Malting barley sells for about double the price per tonne of feed barley (for cattle feed), and sometimes treble or quadruple for some microbreweries that want very high quality barley. The yield per acre is about the same, and the other inputs needed are about the same, except management. There is a very long list of very precise specifications that you must meet to get the best price for malting barley. Because he does almost everything himself, and watches everything very carefully, he can (usually) meet those specs and get that higher price. On a bigger farm, where he would have to hire more workers and contractors, it would be much harder for him to monitor everyone and meet those specs and get that higher price.

Chris:
Yep, farming is a business. (But it's also seen, by farmers I have known, as a calling, or way of life. Most of those guys are as into farming as economists are into economics. They think about farming a lot of the time. When they meet socially, they talk about farming. When they go on holiday they look over the hedges at other people's farms.)

We often have a dichotomy in our minds between idle coupon-clippers of financial assets and active managers of real assets. One is earning capital income and the other is earning labour income. And there is some merit to that dichotomy, even though it's not as sharp as it first appears. (E.g. what about a landowner or house owner who rents out his land or house?) I'm trying out a different perspective. We have two choices: what form do we want to hold our assets in? (those that require little management or those that require a lot); and how much effort do we want to devote to managing our assets?

RSJ: Nope ;) .

[Phil Koop]

"when the strike price of my time falls"

That is probably not the most helpful way to put it. Instead you should say that you have a (real) exchange option, which you will exercise when the value of the asset to be delivered (the opportunity cost of your time) falls below the asset to be received (use of the barn.)

Incidentally, I didn't know that you believe that different sources of income should be taxed at different rates. There might be a blog post in explaining why you believe this.

[Determinant]

Frances makes a good point about mutual funds. Take "Canadian Funds" of any sort really. The Canadian equity market is really quite narrow. It consists of banks, life insurance companies, resource companies, and the rest is a sideshow. Our market is finance and resources.

Let's say you just wanted a slice of the economy. You could buy one of these funds, with 1.5% MER's, or you could cut out the middle man and just buy bank shares. Canadian banks occupy a privileged position in our economy and their returns are reflective of Canada's economic performance as a whole, slightly better actually. This is easy to see if you realize that they are engaged in the business of pushing around our economy's money and then they get a slight premium for being a successful business or owning an economic rent, whichever way you want to see it.

Banks have Dividend Reinvestment Plans. You can buy bank shares from them direct. All you have to do is get at least one share signed into your name. Brokers don't count, they hold shares in "street name" on deposit. Then you just have to sign the enrolment form and you can send a cheque to the bank every month in return for shares. The shares are voted by the Board of Directors.

A Canadian Balanced Fund just layers on more fees for doing the same thing. Why employ a fund manger when you can just employ a bank president?

[rp1]

Asset management is already more important than job income. It has been for over a decade. I don't know why anyone would bother trying to improve their job prospects by getting an education. This decision has arguably never paid off, and now with the extremely inflated costs of education it is a complete suckers bet. Investing at an early age, with money from a high school job, is the way to prosperity. Actually, you could dispense with high school altogether and just give everyone a two month course on inflation and technical analysis at age 13. Working is just for people who want to do it, and the real problem is trying to invent some fiction that will keep them showing up day after day.

[Nick Rowe]

Phil: Yep. Just after I wrote that bit about the strike price of my time, it didn't feel quite right. Your way of putting it sounds better.

The idea that interest income is not really income, and so should not be taxed, is common amongst economists (I think). One of us should explain the reasoning behind it. I've toyed with doing so, but can't yet think up a clear way to express my thoughts on this. Frances would probably be better on this. Except I'm not sure if she believes it. Isn't it a corollary of the Atkinson/Stiglitz theorem (if leisure is separable in the utility function, all consumption goods should be taxed at the same rate, and interest income is just the lower price of future consumption goods)?)

Determinant: yes, as Frances says, mutual funds just means a triple layer of asset management fees. I can never decide if it makes sense, or whether it's better to invest in index funds or re-create your own index fund. The EMH says don't buy actively managed mutual funds. But the contrarian in me says that if everyone else assumes EMH, and buys the index, then EMH is false, and it's better to hold an actively managed fund. (See my old post on the demand and supply of EMH).
http://worthwhile.typepad.com/worthwhile_canadian_initi/2010/01/the-supply-and-demand-for-belief-in-emh.html

rp1: as someone in the education business, I hope you are wrong. But fear you may be at least partly right. But there's more to managing assets than managing financial assets. There are real assets, like land and tractors, underlying those financial assets, and those need people who understand farming to manage them. Technical analysis won't tell you when to cut the barley.

[Determinant]

Index funds are the exception to the rule. They are pass-through vehicles. There are real costs to buying shares. Shares are traded in lots of 100 and generally lots of 1000. If you are a retail investor you can't get past the hurdles of setting up the plan with your $50/week payment. So index funds do that for you. Their MER's are very low, even lower in the US than in Canada so in this particular case you are getting what you pay for and getting a fair deal in services for your payment.

Curiously Benjamin Graham, who was a lecturer at Columbia from the 1930's until the 1950's, mentioned index funds as a concept as early as 1949 in his books. He thought it was an intellectual exercise and quickly moved on. Then John Bogle came along in the 1970's and turned the idea into reality. It's interesting to trace the development of the idea, particularly when standing it next to the Ben Graham devoted friend and follower Warren Buffett.

[Buyer]

Nick:

Chapter one of my micro text states: "In any choice situation, the individual makes the choices that allow him or her to attain the highest possible ranking in his or her preference ordering. This is the theory of self interest." If you believe that we won't manage other people's assets as well as we would manage our own, are you challenging one of the basic tenants of economic theory? Is the Principle-Agent problem a non-issue if you accept that economics is "rational" at its heart? I trust you to manage my assets 'cause you're GOOD at it. I buy things because I'm good at it.

[Nick Rowe]

Determinant: That's a very interesting story about index funds. Normally in economics we assume that the market is efficient, at least in the sense that we economists cannot figure out from our armchairs how market participants could earn higher profits. "If sardine-flavoured ice cream were such a brilliant product, the market would already be producing it; and since it isn't, we must adjust our economic theories to recognise the fact that sardine-flavoured ice cream will fail the test of the market." And here you have a counter-example, where the armchair theorist came up with a product the market hadn't thought up yet.

Buyer: There's a trade-off. My post mentioned the advantages of managing your own assets: avoiding the principal-agent problem. But I ignored the disadvantages. We get other people to do stuff for us (we trade) because of economies of scale and comparative advantage. Most people hire someone else to fix their cars, because they don't have the comparative advantage in fixing cars, and because there are economies of scale in fixing cars. But there's also a principal-agent problem when we trust the mechanic to fix our cars.

[K]

Determinant and Nick:

Index funds are great in principle, but in reality hide a variety of costs. The greatest of these is stock lending revenues, a large chunk of which flow to some third party. Last time I checked Barclay's (as the sole lending counterparty) took fully half of stock lending fees that were generated from iShares stock holdings. Black Rock is probably no better and I would expect that iUnits are even worse than iShares. Awareness of repo/lending markets is essentially zero among retail investors, and about the same among academics. Transparency is abysmal too, so there's little likelihood that we have achieved great efficiency in this area.

[Michael]

I hope you aren't still carrying a mortgage while you place money with fund managers Nick. I find it incredible that any retail fund could beat the return on mortgage pre-payment.

[Nick Rowe]

K: "Awareness of repo/lending markets is essentially zero among retail investors, and about the same among academics."

You are right. I think I am more knowledgeable than most retail investors, and this is totally news to me. And I still don't really understand what you are saying.

Michael: I don't have a mortgage now. I agree that paying off a mortgage is usually a better investment than any financial asset (except maybe inside an RSP, if your current marginal tax rate is much higher than your future marginal tax rate). I confess though I have had both mortgage and stocks in some years in the past. Irrational, I know. Except maybe as a psychological forced savings ploy.

[RSJ]

Nick,

"if leisure is separable in the utility function, all consumption goods should be taxed at the same rate, and interest income is just the lower price of future consumption goods)?)"

I don't understand the connection here -- you can use wage income to defer consumption and interest income to consume. A source of funds is not a use of funds. Now it's one thing to say that only expenditures should be taxed, in which case we would tax neither wage nor interest income. But in aggregate, we know that income = expenditures, so taxing only expenditures is the same as taxing only income. You cannot argue that is inefficient and the other is efficient.

On the other hand, you can also try to argue that you should not tax expenditures on capital goods, only the consumption of capital goods. That's fine, too, accountants would only book the depreciation and or using up of capital goods as expenditures. They would not book the full cost of purchasing the capital good as an expenditure in the period in which the good was purchased.

But none of that reasoning has anything to do with taxing interest income versus wage income.

[Determinant]

I do, and I don't think K's point is really a point. Remember what I said about "street name" holding of shares? That means that the shares that you own are deposited with your broker, or in this case the index fund's broker. The name appearing on the Canadian Depository of Securities is the broker's name, not yours or the fund's. Just like money in a chequing account, securities registered in street name are lent out by the broker at interest. It's what makes Bay Street go round.

So what? So what indeed. This is why the Canadian Investor Protection Fund exists.

K is arguing the same point often levied at chequing accounts: that they should be entitled to a portion of the lending revenue they are used in. That argument doesn't work in banking and it doesn't work in brokering either. It is essentially saying that chartered banking or brokering shouldn't exist.

This is mostly a brokerage issue. The MER for an index fund is quite reasonable for services delivered: transaction costs of following the index, delivery, crediting and reinvestment of dividends, the aggregation of retail investment payments into large wholesale stock holdings, retail administration. Plus sales, though nothing like what a conventional mutual fund pays.

[K]

Nick: As I'm sure you know, short sellers have to pay a fee to borrow the stocks they want to sell. The fee depends on short selling demand. Fees are typically a few tens of basis points but can be as high as many tens of percent per year in cases where there is significant short selling demand. When the expected future fee is high, a stock will trade at a premium since the lending fee acts like an additional dividend. So if you want to earn your rightful return on the stock as its owner, you have to earn the lending fee. If you have a margin account the broker will lend you stock and collect the fee (for himself - not for you). It's a nice little scheme. The ETFs also lend your stock but pay you some of the revenues. Unlike your broker though, as an owner of an ETF, you will be stuck with the loss if the borrowing counterparty doesn't (can't) return the stock. Lending is a credit risky operation and stock lending is barely overcollateralized. The fair value of that credit risk is most certainly not disclosed. Oh yeah, the ETF *itself* also earns a lending fee, also generally earned by the broker. If you want to learn more, google "ETF securities lending."

[K]

I meant: "If you have a margin account the broker will lend *your* stock"

[K]

Determinant: I just saw your comment.

You already pay brokerage fees. There is no competitive efficiency in additional hidden fees that nobody even knows exist. A few months before its bankruptcy, most of the value of GM stock was due to the expected value of lending fees. You could observe this objectively in the forwards market. Many poor suckers were still holding this stock thinking that its positive price implied some non-negligible probability of survival. The stock should have been trading at zero, but instead traded at a premium and earned extremely high lending fees. Anybody who was holding this stock and not earning the lending fee was simply a fool. You want to be the donkey that pulls Bay Street's merry-go-round; knock yourself out. Personally I'll stay away from stocks with high implied future lending rates.

Yes, chequing accounts should earn the risk free rate, and governments should earn the fair value of insuring deposits.

"It is essentially saying that chartered banking or brokering shouldn't exist."

If you say so.

[Nick Rowe]

K: "Nick: As I'm sure you know, short sellers have to pay a fee to borrow the stocks they want to sell."

I didn't really know that. I knew the short seller had to pay the dividends, if any, but you are presumably talking about fees over and above that. It makes sense, since without fees there's nothing in it for the lender. Though I would have thought, absent default risk, that fee would be competed down almost to zero, since the supply of such shares to lend out out to be perfectly elastic at a miniscule fraction of a penny fee, until all are lent out. But it's not something I had ever thought about.

K and Determinant: this is interesting. Still haven't collected my thoughts fully. At a broader level, I would say this confirms the general point of my post. Someone who was actively managing his own assets would collect those fees (from lending his stocks to short sellers) himself. In a world of zero transactions costs and perfect competition it wouldn't matter, since the ETF (or broker) would pass those fees back to you. But a world of zero transactions costs and perfect competition is also a world where Principal-Agent costs are zero. Principal-Agent costs are transactions costs.

RSJ: Assume a 2-good (3 if you include leisure) 1-period model. People earn income from working, which they can spend on either apples or bananas. The government cannot observe leisure, or the wage, but can observe the consumption of apples and bananas. Atkinson-Stiglitz saya that if the Utility function is separable in leisure, the optimal tax system taxes expenditures on apples and bananas at the same rate. For example, if bananas are cheaper than apples, so you can consume 110 bananas for the same price as 100 apples, you don't call those extra 10 bananas "interest income". You don't say that someone who spends all his 100 apples worth of income on bananas has to pay 50 apples tax on his labour income, plus an additional 5 bananas tax on his interest income. You pay exactly the same tax (50 apples=55 bananas worth of tax) regardless of how you spend your income. You don't pay 50 apples+5 bananas=54.5 apples=60 bananas worth of tax if you spend all your income on bananas.

Now, take exactly the same example, only replace "bananas" with "apples next year", and assume r=10%, so it becomes a 1-good, 2-period model.

That's the way I understand it intuitively (why you shouldn't tax interest income). Can any public-finance person add/confirm/deny?

[K]

Nick: Exactly as you say.

"until all are lent out"

Then they become scarce, and the fees can get extremely high.  Especially after a big fall in the stock price, there seems to be an unending supply of knee-jerk contrarian retail investors to prop up the price.  If the market can't reach fair value in the stock price, it finds equilibrium as short sellers drive up the lending fee until the arbitrage opportunity is gone.  Then the only arbitrage that is available is the lending fees that accrue to brokers (and ETF managers) whose clients don't know better and who therefore get stuffed. Lots of these people seem to be under the illusion that they don't have to worry since the invisible hand will somehow look out for them.  But since *none* of the fees are passed on to them, ever, no matter how high they get, this market is in fact the picture of perfect *inefficiency*.

If the forward price (roughly call price - put price + strike) is way below the current stock price and the stock isn't paying big dividends, look out.

[Nick Rowe]

K: when all Nortel shares are lent out, can't I just create a synthetic Nortel share? I get a sheet of paper, and write on it "I promise to pay the owner of this sheet of paper an amount of money equal to NT dividends, plus the market price of an NT share when this sheet of paper is returned to me, signed Nick Rowe".

Why, in other words, do short sellers actually need a real Nortel share to sell? Why isn't the supply of real+synthetic NT shares for short sales perfectly elastic, at a fraction of a penny above the current price, right out to an infinite quantity?

(I expect I ought to know the answer to that question).

[Nick Rowe]

And thanks for educating me, BTW.

[Jacques René Giguère]

Nick:
"Why, in other words, do short sellers actually need a real Nortel share to sell? Why isn't the supply of real+synthetic NT shares for short sales perfectly elastic, at a fraction of a penny above the current price, right out to an infinite quantity?"

That's what the derivative market is all about. Write some paper about other paper, market them to anyone taken in by the mumbo-jumbo and when the premium finally falls,move on to another paper written about a paper written about a paper. Dynamically, the supply of paper is infinitely elastic as is the number of suckers ( oops "dynamically managing your portfolio brokers").
The Street is a circus ruled by Barnum's law ("There is one born every minute") Like the Army where any new 18-year-old volunteer really thinks himself immortal and never wonder why if he is a "reinforcement" the number of men in the field never increases.

[Determinant]

Nick, you just wrote a naked put option.

http://en.wikipedia.org/wiki/Naked_put

Naked puts have substantial risk because you don't own the underlying stock and you haven't limited your risk with respect to the strike price.

[Nick Rowe]

Jacques: yep. But I'm trying to see why the price to borrow shares should be more than a fraction of a penny above zero.

Determinant: agreed. I am making a risky bet. But, it seems to me that I am making exactly the same risky bet if I borrow a real NT share from you, and then sell it short. So why should I pay you a fee to borrow your real NT share, when I can make my own NT share for the cost of the paper and ink?

[Patrick]

"Why, in other words, do short sellers actually need a real Nortel share to sell"

Don't forget that there are practical matters of keeping track of who gets to vote and who gets dividends.

[K]

Every time a share is sold there is some chance it will end up in an account where it is not available for lending.  As more and more shares are short-sold, a greater and greater fraction of them end up in such accounts, and so they become difficult to get hold of.  Furthermore, large institutional holders sometimes decide to stop lending their holdings, thereby causing a "short squeeze" that can suddenly drive up the value of the stock when short sellers buy it back in order deliver it back to their lender.  

What you are describing is something like a forward contract (not a put), but without a fixed maturity date.  I think there are "contracts for difference" (CFD) that accomplish something like what you are describing.  Your contract doesn't get to vote, and it has Nick Rowe credit risk.  But ignoring those details, what price would I pay for your contract?  If I plan to hold it indefinitely, I wouldn't pay the current stock price, S, for it since I don't earn the lending fee from you.  So I'd pay a discount equal to the PV of expected future lending fees.  But if I pay less than S, I could just exercise it right away and walk away with an instant profit.  So I don't think your contract could exist.  Either you wouldn't sell it to me, or I'd exercise it right away.  CFDs do exist, but they offer the same bad deal as stock holding without earning lending fees (plus a lot of other embedded fees, I believe).  Come to think of it, it's a bit like owning interest free money when you could be holding T-Bills.  Shouldn't exist.

Forwards with fixed maturity do, however, exist, and you can trade as many of them as you want.  If there is a large lending fee, they will trade at a big discount to the current stock price, which is fine.  

I think your question may be confusing the "existence of arbitrage
opportunities" with "some people are getting arbitraged".  The market, as seen from the perspective of a hedge fund, is very efficient.  A stock that's going to go bankrupt (or already is bankrupt!) may trade at a value above zero.  But there is no inefficiency since you can't short the stock without paying a hefty borrowing fee (ie. the full amount of the expected short-selling profit).  Alternatively, they could sell a forward contract, but at a price that's *much lower* than the current stock price (by an amount equal to the expected lending fees between now and then).  So there are no arbitrage *opportunities*.  Some people, however, are allowing themselves to be arbitraged by their broker.  Note that this is not an arbitrage trade the broker can instigate at his option.  It is a present to the broker at the option of the owner of the stock.

Which is a long way to saying, yes, the market is perfectly elastic right out to infinity (almost).  But not for you.

"And thanks for educating me, BTW."

That's about 0.1% of what I've learned from you.  Little chance I'll ever get on to repaying the rest.