“Liquidity” is another magic word. I like it rather less than “friction.” In the debate on financial regulatory reform opponents of tight regulations use the word liquidity as a magic spell which, they hope, will make all inconvenient evidence and arguments go away (no links my claim is like saying the Sun emits light). The word is used with different meanings. Sometimes it refers to something which is definitely good, sometimes it refers to the effects of structured finance. This is an equivocation.
It wasn’t always this way. Like “friction,” “liquidity” has been used for purposes other than special pleading and obfuscation. I don’t know the history of the word, so I will start with a definition (which may not be the oldest definition) and explain how the word lost all useful meaning through over use. As with “friction” I think a good practical rule is to demand that anyone who makes an argument including the word “liquidity” be asked to rephrase it without using that word.
Long long post after the jump.
First money is liquid. Money is any asset which can be used to buy anything or pay and debt right now. Other assets are more or less liquid depending on how quickly they can be converted into money and how much that operation costs. One definition of liquidity (the only one which I accept) is the cost of a round trip money to the asset to money divided by the price of the asset.
When we describe the liquidity of an asset as a property of the asset, we are making a very strong totally false assumption. We assume that the proportional cost of the round trip is constant and does not depend on the amount of the asset purchased then sold. Of course this is nonsense. For extremely small purchases, the cost of a round trip is very high. Much more importantly the round trip costs of extremely large purchases can be much higher than the round trip costs of medium size purchases. An extremely large buy order drives prices up and an extremely large sell order drives prices down. It is this problem and no other which is being discussed basically every time a financier uses the word liquidity. They call an asset liquid when huge positions can be taken and unwound with low round trip costs. This sort of liquidity – large transaction liquidity – is not directly valuable to me or to 99.9% of people.
The main point of this post is to object to the equation of round trip costs for medium size round trips and huge round trips. However, the abuse of language is much worse. A third meaning of liquidity is the money supply. This is the sense in which Central banks are said to inject and remove liquidity via open market operations. Here “liquidity” means “liquid assets.” I don’t think this abuse of English is dangerous. It is just very mildly irritating.
However, and much more importantly, “liquidity” is also used to mean free cash flow. It is in this sense that people distinguish between an illiquid firm (which can be profitably rescued with a loan) and an insolvent firm (which can’t). The claim is that the problem for an illiquid firm is that its assets can’t be converted into money.
This is clearly a false statement. Nowadays all assets can be sold. Even in the past, firms could issue new shares. The problem for an illiquid firm (with this 4th meaning of liquidity) is that no one wants to buy its assets. The claim of the manager that her firm is illiquid is a claim that the asset prices are different from fundamental values.
The assets can be sold (as shares of the firm if need be) but people would not pay enough to cover the firm’s liabilities. This is not because the instrument called the public offering of shares doesn’t exist. It is because they don’t believe the assets are worth that much. The claim of illiquidity not insolvency always is the claim that investors are wrong about fundamental values. It hasn’t been a statement about available financial instruments since the invention of the joint stock limited liability company.
It doesn’t matter if the firm’s effort to liquidate its assets would drive prices down or if prices are already low. What matters is that the firm can’t get enough money for its assets to pay its debts.
For the purposes of this post, the point is that, other things equal, we would like firms to have a huge free cash flow. Using “liquidity” to refer to free cash flow gives the word positive connotations. The implicit suggestion is that market institutions such that huge round trips have low proportional costs will improve the cash flow of all firms. This suggestion comes close to being an explicit argument when the same person describes the recent crisis as a liquidity crisis and says that tighter regulations will reduce liquidity.
A fifth meaning of “liquidity” is “market thickness” or “trading volume.” I’m fairly sure the word originally described assets. Now it is often used to modify markets so thick markets are called “liquid markets.” This is actually fairly important, since market thickness is key to making the round trip costs of taking and unwinding a large position low. If I own 100 shares of IBM, the amount of money I can get for them depends almost not at all on trading volume the day I sell. A minuscule fraction of current market volume would be enough that I don’t drive the price down by dumping 100 shares.
The identification of round trip costs for huge and medium round trips leads many people to consider high trading volume socially desirable. Just try to imagine an explanation of why high trading volume is useful to someone who buys corporate bonds to save for her retirement or pay her children’s college tuition or buy a house. It is just assumed that market participants frequently take and unwind huge positions. The implicit assumption is that we should regulate so that the properties of the market please active traders. Most people do not believe that active traders are socially useful. The ambiguity of the word liquidity enables them to argue that the market should be designed for their convenience without their having to defend their claim to be socially useful.
Now an analogy – poker. The thing that smart professional poker players want most is not so smart poker players (fish, suckers etc). They want people who bet incorrectly so that they can take our money (they haven’t gotten any from me). Now, what about the smart traders who want liquid markets? Their view of the way markets should be is that, when they decide that an asset is underpriced, they can buy a whole lot of it at that too low price and, when they decide that an asset is overpriced, they can sell a whole lot of it at that too high price. I can see why they think this is good for them (I think they are often wrong, because they don’t have rational expectations). I can’t see why we should feel obliged to supply them with plenty of fish.
OK an example. “The RMBS based CDO market became very illiquid in 2008.” This sounds like a safe claim. The fact is that the volume of trading of RMBS CDOs collapsed so the market became thin. A market can be thin if everyone agrees on the fundamental value of assets and is pleased with their current portfolio. This was not the case in 2008. Many firms were very eager to get the CDOs off their books and no one had a clue what they were worth. A market can be thin as it is perceived to be very risky so everyone likes to hold zero of the asset. This was not the case in 2008. Net holdings of RMBS base CDOs were not zero. Many firms and individuals owned them and felt that they were bearing risk. They would have liked to reduce their holdings.
I think it is clear why the market was thin. The market clearing price was much lower than it had recently been. Very few wanted to buy at any price higher than this market clearing price, but the owners of the CDOs weren’t willing to sell at the market price, because then they would have had to book the losses. The market seized up, because it was necessary for it to seize up to neutralize mark to market accounting. Here as in the case of the allegedly illiquid firm the problem isn’t that the assets were illiquid, the problem was that their market price was so low that firms couldn’t cover their liabilities by selling the assets.
Having no respect at all for the efficient markets hypothesis, I am perfectly prepared to believe that this was an irrational panic and that many firms could profitably be saved by an emergency loan. It is not very bold to type that now, but at the time, I thought that the cost of a well designed TARP to the Treasury would be huge and negative – that the US Treasury had a huge gigantic profit opportunity.
This opportunity was largely wasted. My understanding is that the Treasury now guesses that it plus the Fed will make money saving banks (I count making good AIG fp obligations part of saving the banks). Not enough to cover the losses from saving GM Chrystler, Fannie Mae and Freddie Mac but quite enough to make it impossible for any semi-sane person to take the EMH seriously (of course I think no semi-sane person ever took the EMH seriously and I strongly suspect that Eugene Fama is semi-sane).
Well here we are finally at the end of an overlong post. My conclusion is that the debate on financial reform is seriously endangered because of a gross equivocation. It is insinuated that so long as trading volume is high, all firms will have an excellent cash flow. It is almost stated that extremely high trading volume is useful to savers and small investors. It is assumed that the problem in 2008 was that markets stopped working in a way pleasing to active traders who think that they can beat the market.
By the definition given, the real estate crisis is a liquidity problem, people can not sell at a price they want to so they are illiquid. (Of course real estate is a post child for an illiquid asset as no two pieces of real estate are totally identical) .
If you look at liquidity stocks are most liquid, with sp 500 being best, options less, fixed income being less liquid to illiquid. Recall that most of the money wall street makes is on fixed income. One measure of liquidity that can be used is the bid/ask spread, the smaller the spread the more liquid the market.
To the point if you invest in mutual funds then the issue is what is the trading cost to the fund of its maybe large purchases and sales (in particular actively managed funds, index funds have this problem far less) The Partnership (history of Goldman Sachs) says that block trading came about because of the rise of the institutional investor who moves 100k share blocks not the 100 share blocks of the individual. Goldman would offer the instutional investor a price for the whole block and then resell it saving the institution from lots of little trades (and in the day this started huge commissions being pre decontrol)
Look if I have a sock and I’d like to sell it for a million dollars it would not be a liquidity problem. This is a sanity problem. You can’t possibly define a liquidity problem as occuring when “can not sell at a price they want to so they are illiquid.” You write “by the definition given.” Given by whom ? Pangloss or the tooth fairy. People can’t always get whatever they want. Trying to design financial markets so people can always sell things for the price they want is not a good idea.
By the definition given by me, houses are very illiquid. The transactions costs of buying and selling a house are huge. If you bought a house today and sold it tomorrow the operation would cost you a huge amount of money. This is not a description of the real estate crisis. Things have always been that way, and real estate is not always in crisis.
YOu assume that high trading volume is good writing “sp 500 being best.” If I am saving for retirement and will buy an asset now and maybe sell it once, why is liquidity so important to me ? How much of my return is lost to one bid ask spread ?
Note that liquidity as used in the debate definitely does not mean bid ask spreads. It refers to the price pressure effects of huge trades. This is unrelated to brokerage fees and barely related to bid-ask spreads. It has nothing to do with individual investors of reasonable means. It also has nothing to do with say Warren Buffet who trades rarely. I’d say it has nothing to do with a well managed pension fund which buys and holds the S&P 500 (I recall a paper which noted that pension funds on average did significantly worse than the S&P 500 while holding an average S&P beta of 1).
f I am saving for retirement and will buy an asset now and maybe sell it once, why is liquidity so important to me ? How much of my return is lost to one bid ask spread ?
Maybe you’re a little bit different, but most people who are saving for retirement buy an asset this week, and then more in two weeks, and more two weeks after that, etc. Most “semi-sane” financial advisors then tell them to rebalance the portfolio towards target asset allocations along the way usually no less frequently than annually, requiring both buying and selling. When they are in retirement, they sell such assets presumably at least once per year.
To Lyle’s point, and I believe you’ve captured it somewhere in your definitions, one way of describing (defining?) liquidity is the profit available to the intermediary between the two trading principals. Since (putting words in your mouth) you think principals have limited social usefulness, reducing the profits available to them would seem to be socially desireable. Fixed Income, Commodities, Currencies, Derivatives, are hugely profitable businesses for the market-makers, equities much less so. While these markets are generally deep, the assets that are being bought and sold are typiclaly much more bespoke than the common equity of a large well known company. While it’s not necessary to have high trading volume to reduce the profits available to the principal, it certainly helps.
One point of disagreement. Suppose I own a house, that I can sell for $300,000 if I have a month to sell it. But I can only sell it for $200,000 if I have one day to sell it. And I have a mortgage for $250,000 that is due today. I would say I face a liquidity crisis, but not a solvency crisis. I agree there’s no precise dividing line (why 1 month?), but there does seem to be a useful distinction here.
On your definition of liquidity. It’s fine as far as it goes, but it does seem to miss something important. What is the correlation between my wanting to sell an asset quickly and lots of other people wanting to sell that same asset quickly (because we all need cash at the same time)? If that correlation is high, I would want to say the asset is illiquid.
I think that most people do not trade systematically.
we don’t agree about buying assets on more than one occasion. I said I would buy the asset once and maybe sell it or maybe hold it till maturity. I am talking about say the bond I buy today. Buying another bond tomorrow is neither buying nor selling the bond I bought today.
If you are losing any significant income to fees or bid ask spreads following your financial advisor’s advice to rebalance, then, I think you are being cheated. If you can present any evidence at all that a portfolio balanced as an advisor advises this year outperforms (in return or more likely risk) a portfolio balanced as that advisor advised last year, I would be very interested.
Even if you believe, as I do, that financial advisors don’t know anything useful which I don’t know, you will be unsurprised that their advice is not “hold the portfolio” year after year. If they never say anything new, you might think you don’t need them. I think you don’t need them, and that to hide this fact, they come up with the bogus claim that they can tell how to balance a portfolio so well that it is worth fees and bid asked spreads to follow their advice on how to react to new information. What would they advise if the best strategy for the investor was buy once and hold ?
I might add that these ideas are not heterodox in the academy.
Yes you are totally right. In my mind, I was assuming the cost is the roundtrip cost if you insist on buying within one trading day and selling within one trading day. Really for liquidation of a given position, liquidity is still a function of how soon you require the money. So liquidity for an asset is a function of two variables — the volume of the transaction and the required speed. I guess (it is just a guess) that what often matters is the amount to be sold per day so more rush and more to sell are similar.
I guess I left out the “I’m assuming you mean convert to money within the trading day.” Since you think a month is reasonable, this sure didn’t go without saying.
I wouldn’t call the fact that many want to sell when I want to sell liquidity. I’d say that the appropriate estimate of risk is minus the correlation of returns and my marginal utility of money (of consumption for an individual). This is the officially accepted standard model in mathematical finance. Poor returns exactly when I need money are high consumption beta. Now I realize that real world investors (including investment banks own account) often use the variance of returns on their portfolio as a measure of the risk they are bearing. This is just a mistake. It is an error along the line of assuming that the risk of an asset is its own variance and not its market beta — that is own variance is the wrong measure, market beta is better but still not right , consumption beta is the correct measure.
‘Liquidity’ is the ability to realize the full utility of an asset. In abstract terms, Let ‘A’ be the asset, let ‘U’ be the utility function, and transformation function ‘T’ is the process of transforming one asset to another.
Liquidity = Log (U(T(A)) – U(A))
Note that the Utility function is subjective.
Money is not necessarily liquid. During period of hyperinflation, a currency becomes illiquid. Zimbabwe was a good example, there was a period when cooking oil and bread is more ‘liquid’ than cash!!
rdan – everyone who contributes to a 401(k) plan trades systematically. Outside of defined contribution plans, I would agree with your statement.
If you can present any evidence at all that a portfolio balanced as an advisor advises this year outperforms (in return or more likely risk) a portfolio balanced as that advisor advised last year, I would be very interested.
While I could be wrong, believe that a substantial portion of portfolio returns has more to do with getting the “right” asset allocation than which manager or idiosyncratic security that fills the bucket of that particular asset allocation. Thus if your target portfolio is 60% equities and 40% bonds and equities decline 30% in a year where bonds decline 10%, you, in theory, should rebalance your year-end weightings away from bonds and into stocks. The portfolio recommendation would be the same, but differential market performance would require rebalance.
I want to follow up on this, if I may. You are helping me organise my thoughts.
I agree with what you say about risk being minus correlation between returns and MU of C. What I am trying to do is to try to think about liquidity within that same sort of framework.
For example, suppose there is an asset that is perfectly liquid, except on Mondays. It doesn’t trade on Mondays, so you can’t sell it at any positive price. If my MU of C is always low on Mondays, that’s not a problem for me. I will never want to sell it on a Monday anyway. But if my MU of C is sometimes unpredictably high on Mondays, that’s a problem. The asset has strongly negative returns if sold on a Monday (it’s as if the price of the asset were zero on Mondays). What matters is the correlation between the returns on the asset and my MU of C.
So an analysis of the liquidity of the asset, and whether “illiquidity” matters, ends up in very much the same place as an analysis of the riskiness of an asset, using the standard proper definition of risk.
Next step. If I need cash, and I sell a large amount of an asset quickly, to make the round trip, I may push the price down. The more I sell, the more I push the price down. That’s part of what it means to say the asset is illiquid. But if other people are doing the same thing at the same time, making the same round trip, collectively we push the price down more. So the correlation between my need for cash (MU of C) and others’ need for cash matters. So illiquidity is starting to look a lot like risk, when we analyse it.
My head is not clear on this. I am trying to organise my thoughts. I can think of examples (like a bad harvest in an agricultural economy) where the correlation between returns on an asset and MU of C have nothing to do with liquidity, only risk. But there are other examples, like everyone rushing for the exits at once, where it seems that the correlation between returns and MU of C is capturing something about illiquidity.
***Nowadays all assets can be sold. Even in the past, firms could issue new shares.***
Of course, but issuing new shares is not exactly the same as selling assets. You and I own a company — Fly By Night Ltd. FBL has invested heavily in Greek Government bonds. We’d like very much to sell said bonds, but we find when we talk to brokers that not only is the listed price very low, at the quantity we wish to sell, there are no buyers. None. Nada.
We can issue new stock for FBL. We may even be able to peddle it to Widows, Orphans, bankers and other naive buyers. But that isn’t the same as selling the bonds. The bonds still are in our possession. We have merely diluted the ownership. The bonds themselves remain in FBN’s account or safe deposit box or under out matresses. We need a word to describe the bond’s condition of having some unknown value, but no current, active, market. “illiquid” seems apt.
t is assumed that the problem in 2008 was that markets stopped working in a way pleasing to active traders who think that they can beat the market.
I never thought this. I always thought that the market went down because the underlying commodities being traded lost their value and that some financial instruments required fixed payments that made some firms insolvent that could not pay them.
Of course since you are paying the advisor you expect some action from the advisor, for buy and hold why bother to pay the advisor? (Catch 22 here). Boogle says essentially you can’t beat the market in the long term, in fact with fees you can’t even keep up, so look for the smallest fees on the broadest securities. This does require humility as you admit you are not up to beating the market long term.
Liquidity can also be viewed as the haircut required to borrow for the asset. You can borrow against treasuries with a low haircut — less than 10%. That means that with $100, you can buy $1000 of treasuries. As a result, if you think the treasuries are mis-priced, then you will be able to afford to buy them — demand can expand quickly when the price falls below the expected value.
So the haircut required can be viewed as a measure of liquidity. And for securities that can be readily used as collateral with low haircuts, you can argue that if you cannot sell the security, it is not because the security is illiquid, but because your price is too high.
A better definition would be the slope of the demand curve with respect to quantity issued, or the p.e.d., except that you would need to replace “price” by “deviation of bid price from expected return”. The advantage of the haircut approach is that it can be explictly measured.
With this definition, the Fed was attempting to add liquidity by allowing dealers and banks to borrow against non-treasury collateral via the lending facilities.
…but this still leaves the possibility that the security is not selling because people realize that they have no idea what it is they are buying — even if the instrument qualifies as collateral by a CB lending program.
I don’t like to think of that as a liquidity crisis — a “transparency crisis”? Fraud?