Saturday, February 14, 2009

On banking

Like many people, I have found myself unusually interested in topics like economics and banking lately. I've watched much of my net worth go up in smoke as housing and stock prices have plummeted, I've watched friends and family struggle through lost jobs, and I realized that I really had little idea of what is causing all of this.

So I've been on a crash-course lately. One of the first things I've read on the subject still remains perhaps my major candidate: our "financial system", by which I mean our currency system, our banking system, and the federal reserve system that is the engine for both of of those.

Let me be clear: I'm not *just* talking about the current economic crisis. I'm talking about the repeated *pattern* of such crises, from the dotcom bubble, back to the recession of the late 80s/early 90s, and going back further to the economic problems of the late 70s and early 80s (i remember mortgage rates in the high teens, though I was too young to have one myself).

And I'm really not talking just about the US: it's clear that many if not most other countries are experiencing similar boom and bust cycles.

So why currency and banking? For one, it is my understanding that most countries now have the same basic system as the US, more or less the "federal reserve system". In particular, the bugaboos that I see are A) fiat currency, that is, currency not backed by something of intrinsic value (like gold), and B) fractional reserve banking, that is, the policy by which a bank can accept money from depositors based on the promise that *at any time, that depositor can have his money back*, and then they promptly turn around and loan the vast majority of it out (the exact amount that they have to keep is the "fraction" that they have to keep in "reserve").

This post in the NYTimes freakonomics blog is pretty typical of the reasoned, relatively politically impartial explanations and discussions I have seen about the economic crisis:

http://freakonomics.blogs.nytimes.com/2008/10/15/everything-you-need-to-know-about-the-financial-crisis-a-guest-post-by-diamond-and-kashyap/

Until this week, it appeared to be increasingly possible that there could be widespread bank failures that would impair the lending capacity of the banking system for a long time. We agree with the words of Ben Bernanke this week: “As in all past crises, at the root of the problem is a loss of confidence by investors and the public in the strength of key financial institutions and markets …”

Banks are vulnerable to loss of confidence because they rely on short-term funding.
Of course there's a lot more in the post, but I'm struck time and again by this same sort of message: the underlying problem seems to be that banks could fail, and the driving of that possible failure is "loss of confidence".

Take a step back: does it sound like a good idea to build the entire concept of a nation's prosperity and longevity on something as ephemeral as "confidence"?

Isn't there a small voice in the back of your mind that says "perhaps it should be based on solid principles that *can't* fail instead of 'confidence'"?

And to be really cynical: don't we have a special, pejorative term for businessman that rely on "confidence", that is, "confidence man"? And isn't that term reserved for a con-artist who only gets away with his cons because of the confidence his marks have in him?

Shouldn't, in other words, our *legitimate* banking system be based on something other than what con-artists based their success on? Or, put another way: if both con-artists and bankers depend on "confidence" for their success, are they really *different*?

These questions aren't just rhetorical when you dig into the details: there's a disturbing similarity between fractional reserve banking as it is practiced today and so-called "Ponzi schemes" as practiced by con-artists (such as the currently infamous Bernard Madoff).

Let's review Ponzi schemes (or more generally pyramid schemes): the con-artist hangs out a shingle advertising himself as a place for investors to invest their money at good rates of return. Investors originally come to him because of some sort of confidence in the con-artist: he has a good story to sell, he's a good salesman, he prints up glossy brochures, whatever. He tells the investors that they will get a good rate of return, and that they can take their money out of the scheme at any time. Original investors give him money, and he duly reports that they have made additional money. If the investors are convinced that their money is safe and growing in the scheme, generally they do not make withdrawals. For the few investors that do make withdrawals, the con-artist simply gives them back money he has taken from them and other investors. Often, the con-artist uses those withdrawers as "proof" to other current and potential investors that the scheme is legitimate, because look, that person got his money back and then some! Do you see the pattern? As long as enough investors have "confidence" in the scheme so that they leave the money in the scheme and don't withdraw it, the con-artist can continue taking money and spending it on himself without ever actually investing the money. The scheme can continue for as long as the confidence is sufficient to bring in more money than is being withdrawn, and in the meantime, the perception on the part of the investors is that things are going great, because on paper, they appear to be getting wealthy fast.

Of course, eventually what happens is that something happens and people start to lose confidence and withdraw their money, and the deception and con are laid bare, and many people lose a *lot* of money.

But wait! What if there were a *network* of con-artists, each running their own Ponzi schemes, that cooperated to cover each other's "runs"? In other words, if one scheme gets a "run", the other schemes step in and temporarily loan him the money to cover his run - in a short-term, for-interest loan naturally - until his investors get their confidence back? Once his scheme is flowing again, he can pay back those loans, and the crisis is averted for the whole collective of ponzi-schemers.

Except: can't that also fail if there is a *massive* loss of confidence in *all* ponzi schemes? At this point, the schemers can't borrow from anyone because they are all in trouble, short-term loans from other schemers dry up because none of them has the extra cash to spare, and what can happen is a domino effect as one after another scheme is laid bare and goes insolvent.

Unless, of course, there's a "lender of last resort": an entity that can print infinte money and that can step into the void and prop up the entire Ponzi-scheme industry when there is a massive, industry-wide "lack of confidence".

I admit that similarity is not "proof", but it's at least a warning flag: doesn't the above sound a *lot* like our current banking system, with "ponzi scheme" replaced by "bank", "con-artist" replaced by "banker", and "lender of last resort" being played by the Federal Reserve?

I am not at all convinced that that parallel is not accurate.

I suppose one legitimate question that I don't think I'll attempt to answer in this post is "well, so what? *If* it's the case that the banking scheme as currently constructed is at its roots a ponzi scheme *but it still provides great value to society* because, say, it "greases economic wheels" in a way that a less ephemeral system might not, then merely noting that our system *is* a ponzi system may not be enough to justify changing that. I'm skeptical that I could be convinced of that - again, look at the freakanomics blog and look how much it talks about the need to mitigate *exactly* the kinds of problems a network of ponzi schemes would have to mitigate to conclude (at least tentatively) that if we didn't have that scheme, we wouldn't have those problems to mitigate in the first place, and thus arguably wouldn't be having these periods of booms and busts - but I'll leave that on the table for some later day.

Meanwhile, let me return to this issue of banking. From the day I was a kid, I've been told and "understood" that when I made a deposit in a bank, it didn't actually *stay* in the bank, it was lent out to someone else. And I bought that (I believe I've blogged about the scene in "It's a Wonderful Life" where Jimmy Stewart sells this idea lock stock and barrel).

But the more I think about it, the more something doesn't smell right. For one, symmetry is usually a good sign. When I take out a loan from a bank - that is, when they give me money on the promise that they'll get their money back later - I do not take it on the basis that the bank can come and get their money any time they want. In fact, I have a very measure timeline over which they will get their money back: they get a certain amount once a month, for a fixed period of time. Because I know exactly what that schedule of repayment is, I can safely manage spending that money in a way in which I will never be "surprised" by a "run" on the money I've borrowed from the bank. The bank does not need to have "confidence" that I have the money: they have a schedule saying exactly when they will get the money paid back, and in the event that I fail to keep to that schedule, they have the collateral that I've put up against that loan, e.g. my house in a mortgage loan.

But the lack of symmetry is startling. In terms of first principles, there is no difference between a bank giving me money for me to eventually give back, and me giving the bank money for them to eventually give back to me. And yet, the terms are completely different. When I give the bank money - at least in a standard deposit - the bank does not have a schedule on which to give it bank to me. Instead, it is as if they took a loan from me that I can insist at any time I want the entire balance back, instantly. Who would take a loan like that? It seems clear to me that the same reasons an individual doesn't take a loan like that - basically, because it makes the money unusable - are the same reasons such a loan would have no value to a bank. And yet, they take it... just like the con-artist takes the money into his Ponzi scheme on the false promise that all of his investors can have their money back in total any time they want. The second that bank actually *does* something with that money, they've made it so that they cannot keep their promise to all of the people they have taken a loan from.

Doesn't that sound like it could be a root cause of a lot of ills? We have people signing contracts that they know they cannot keep, and this is the fundamental basis of the federal reserve system. It *is* a ponzi scheme, and it is therefore no wonder that the scheme depends on the "confidence" of the people who have given their money into the scheme in the mistaken belief that the schemers can actually keep the contract they made with them.

And all seems great while confidence is high, as in all Ponzi schemes: the depositors leave their money in the scheme and see great numbers on paper and feel wealthy and may even go out and spend more on consumer items and other immediate consumption than they normally would, and that money goes into the hands of the people who sold those products and they feel confident and continue to spend the money, all because on *paper*, they have a lot of wealth stored in the ponzi scheme.

That spending/consumption has a reverberation in the system, though: it represents withdrawals from the ponzi schemes. Ponzi schemes can survive a certain amount of withdrawal - as noted, they pay withdrawals from others' deposits - but they can be a victim of their own success: as people see a *lot* of wealth on paper, they feel more comfortable withdrawing that money *for the consumption we just mentioned*. As that spending continues, the Ponzi schemes start to get closer and closer to the breaking point. They are of course still exuding confidence: they have to, that's what their scheme is based on. But ironically, too much exuded confidence continues to prompt depositors to take their paper wealth and spend it, bring the schemes to the popoint of ruin. The first schemes to hit this point will survive by the trick of borrowing from other schemes, but if the depositor's behavior is widespread, eventually they are all going to hit the edge, and the whole system is going to come crashing down.

Note that this crash can also be exacerbated or caused by externalities that cause a lack of confidence, e.g. 9/11, but the dynamics here seem to indicate that even without the externality, the system is bound to run into boom/bust cycles.

I'm not in any way saying that this is a thoroughly educated economics proof, but it sure does pass at least the "laugh test" for me. It is plausible, it is explanatory, it is predictive, and those predictions match the general pattern of what we've seen historically, as well as even the descriptions of what we are seeing in the responses to the current crisis: it sounds like a group of ponzi schemers trying to figure out how to keep their scheme afloat. Look at the specifics: the banks are "trying to get credit flowing again", or in ponzi-speak, the ponzi schemers are trying to get enough above water that they can start covering each other's "runs" again. In fact, the term "run" is explicitly used in the discussions of the banking industry's problems!

So if this is true, then how to fix it? I think it starts right at the bottom, in the relationship that banks have with depositors: no one should ever agree to a contract that they can't keep. Just as a person taking a loan from a bank does not falsely promise to pay back that loan whenever the bank wants it and backs up their loan with something tangible that will go to the bank if they can't otherwise pay the bank, a bank should only make contracts with its depositors that it can keep. If it takes in a depositor's money based on the notion "you get your money whenever you want", then it needs to *keep that money*. No fractional reserve banking *on that kind of a deposit*. True, the banking industry would need to change a little, but that's kind of the whole point here, right, if it's the banking system that's at the heart of the boom/bust cycles, then we *want* to change the banking system. If a bank is just going to keep your money in a vault - literally just be a safe place to store your money, the way it is adverstised - they inevitably you are going to have to pay them for that, just as you pay the UStorage place to keep your stuff in storage. The free market will eventually work out good compromises and rates here.

Where will banks get money to loan out, if not from deposits? Basically, they need to "borrow" money from individuals in the symmetric way that individuals borrow money from banks. One way in which this concept exists today is in the form of "certificates of deposity". A CD is analagous to a loan: the bank does *not* promise that the money can be had at any time in its entirety, but instead agrees to a fixed schedule for repayment, with some extra interest on top. Because the bank knows that this money can't be "run", it can figure out how to loan that money out *and still pay the money back on the schedule that it agreed to*. Think about it: when you get a loan, say, it is not uncommon to think "I'll need to pay back this loan starting next month, so I'll put a bit of the loan aside to make the first few payments; I'll then use the rest to buy the car I wanted, and I know from my analysis of my cashflow situation that I'll be able to start making the payments from my salary starting in a couple of months." That is, you can spend the money from the loan while still carefully planning on how to pay back the loan. With a CD or other "loan" like instrument, a bank can do exactly the same thing. They will know their payback schedule and the payback schedules of the loans *they* have made, and it's an easy accounting exercise to make sure that they can meet their obligations. Note that the bank is not in danger even if their outstanding loans go belly up, as long as they did their homework in evaluting the collateral: if the loan defauls, they take the collateral and convert it into the cash they need to pay back their CD holders. To be sure, such a bank can *purchase* insurance to cover the very small chance that they lose money on a loan and cannot quite cover a CD obligation, but as that should be very rare and a very small amount of money, such insurace such be inexpensive relatively speaking, and of course the bank should account for that cost in the interest it charges on its outgoing loans.

No "confidence" is needed in such a system. There is no "ponzi scheme" element to such a scheme. It works because it is a series of contracts that are actually intended and guaranteed. There cannont be a "run" on such banks.

It's pretty clear that individuals would participate. How many would choose to just keep their money in a bank for a fee? Probably not too many: they know that they can make interest on that money in a CD, so they'd probably choose to put most of their excess into CDs, leaving enough in the bank to cover immediate cashflow. Thus, most of the money would still be available to the banks for loaning. But that money would not require *confidence* to loan back out, it would simply require *accounting*: there's no risk that the holders of CDs can "run the bank", and so banks can lend it out with clarity about their obligations and determinism about what they do and will hold and when.

So in my naive, non trained-economists view, this appears to solve the problems with "runs" on banks and "confidence" while still allowing capital to flow throughout the economy. It allows banks to make money, but arguably less money than they currently make, for the benefit of reducing the "boom/bust" cycles that seem to naturally flow from the ponzi-scheme setup that most of all current nations currently have. And it eliminates the need for a "lender of last resort" and thus the currency manipulation on the part of the Federal Reserve or its equivalent in other countries that seems to contribute to many woes (admittedly, I haven't gone into that much here, but this post is long enough).