One theory is that Mt. Gox became fractional reserve some time in the past, either by losing bitcoins or by spending them. Then they were unable to replenish their supply, so it was a matter of time before a bank run.
The reason I tend to believe this scenario is because it's completely consistent with their behavior. I watched them very closely, and there seemed to be no rhyme or reason for their behavior. That is, unless they were missing everyone's bitcoin for some reason. Then their behavior made perfect sense.
In that scenario, Mt. Gox would have knowingly traded non-existent Bitcoins for far, far longer than two weeks.
EDIT: I should mention that there's still no evidence whatsoever that malleability somehow led to the loss of >500,000 BTC.
> One theory is that Mt. Gox became fractional reserve some time in the past, either by losing bitcoins or by spending them.
Maybe pedantic, but many people have been using fractional reserve wrt Mt Gox lately. The usual meaning of that term: A bank will loan out deposits, reserving a fraction for withdrawals. However, a key point is that the bank holds collateral against the loan, and that collateral has a fair-market value, so the balance sheet is still positive. (A major problem in the housing meltdown was that the value of the collateral dropped, making many banks technically insolvent.)
What Mt Gox did is take money from depositors and either lose or spend them. That's just either bad business (if they lost them) or fraud (if they spent them). Calling it "fractional reserve" gives it an air of legitimacy that they really do not deserve.
Edit: I did not intend to start a discussion on the finer points of bank accounting. The major point is: To my knowledge, Gox wasn't trying to make loans with money that deposited with them, which is what a fractional reserve business is.
> Maybe pedantic, but many people have been using fractional reserve wrt Mt Gox lately. The usual meaning of that term: A bank will loan out deposits, reserving a fraction for withdrawals. However, a key point is that the bank holds collateral against the loan, and that collateral has a fair-market value, so the balance sheet is still positive.
No, the balance sheet is still positive because the debt owed to the bank is an asset of the bank. This is independent of whether the debt is secured by collateral.
(Of course, even a risky loan that is unsecured by collateral is a very different thing than simply having deposits stolen, so, there is a good point that while Mt.Gox surely had less-than-full reserves, it was doing something very different than fractional reserve banking, even assuming that Mt. Gox's own explanations are correct.)
> No, the balance sheet is still positive because the debt owed to the bank is an asset of the bank. This is independent of whether the debt is secured by collateral.
Not always true. For non-recourse loans, the value never exceeds the collateral. When the collateral gets written down, so does the asset.
Or, put alternately, it taints actual fractional reserve banking (which has worked pretty well for hundreds of years) with a sense of being the same kind of shady business practice as this. (Which may be the point of the comparison, given the ideological bent of many Bitcoin advocates.)
i don't get how you can loan bitcoins in any other way other than full reserve - unless you somehow put trust into the bank (which you have to for fiat money, but for bitcoins, you can verify and so don't have to trust credit/notes offered by the lender).
Same way you don't get back the exact serial-numbered dollars you loan (deposit) to a bank. It's not the _identity_ of the bitcoin which matters, it's the value.
If you want a lockbox, get a lockbox. If you want a deposit account, you're storing and retrieving fully fungible and interchangeable entities.
And it's worth remembering that banks aren't the only institutions that make loans. There are a lot of businesses that extend credit (post-paid services, invoicing with net 60 terms, etc), which has the same exact macroeconomic effect.
Probably only the sort of fractional-reserve banking in which the banker lends out for a year or 30 deposits accepted under the promise to return them upon demand. That's the sort that can have bank runs. If depositors agree that their money is locked up for a while, no problem.
Huh? This isn't true. Banks loan without collateral all the time, such as via credit cards or unsecured lines of credit. This is why fractional reserve banking increases the money supply. Banks are allowed to literally invent money out of thin air, so long as their invented money is within the multiple required by law for cash they have on hand.
Mt. Gox was doing (attempting) the same thing our traditional banks do in that sense. They were increasing the bitcoin supply using an analogous scheme.
> Banks loan without collateral all the time, such as via credit cards or unsecured lines of credit.
Yes, I was avoiding that can of worms. Banks do unsecured lending, there is still an asset entry to offset it so that the books remain positive. Armies of regulators and accountants and volumes of laws in effect here.
But the salient point: A fractional reserve business consciously makes loans with an expectation of being paid back. To my knowledge, Gox was not trying run a fractional reserve business, and the term is being misapplied.
> This is why fractional reserve banking increases the money supply.
It increases a money supply, not the money supply. It does not increase M0 (and there was a time that banks were allowed to do just that.)
Note that in a secured loan, the security is rarely for the full amount of the loan (mortgages being a notable exception). Usually it covers only a fraction of it, but it's there to increase the borrower's "skin in the game". Rules also differ on whether or not the security itself is acceptable as discharging a loan's obligations (e.g., in Spanish real estate lending, if the property itself isn't sufficiently highly valued to pay off the loan balance, the borrower _remains_ on the hook for the balance, this is usually not the case in the US, where most mortgages are non-recourse loans.). See also "deficiency judgement".
In both secured and unsecured loans, the debt itself is the security, with a portion of the interest being attributable to the risk (default) component of the loan.
As for money supply, bank reserves are included in M2 which is used for inflation calculations, so I'd argue that yes, fractional reserve lending does increase money supply, as commonly used.
"Banks loan without collateral all the time, such as via credit cards or unsecured lines of credit."
It's true that there's no collateral, but there is a corresponding asset - the loan itself.
"This is why fractional reserve banking increases the money supply."
Fractional reserve banking would increase the money supply even if banks restricted themselves to fully secured lending. Bank deposit accounts act a lot more like money than does a mortgage.
i want to imagine uncollateralized load like this:
i'm a bank with 0 dollars in any asset, but 0 debt obligations. So my networth is 0 right now.
You come along to borrow off me $100. Now i have an asset of $100(the debt which you have to pay back + interest), but as soon as you spend your money, i also have a debt of $100.
Now isn't this a good way to make money from nothing?
If I have zero dollars, how did I have anything to loan you? I can put numbers on my ledger, but you won't be able to withdraw it - that's not fractional reserve, that's zero-reserve.
On the other hand, I've never objected to the notion that fractional reserve creates money - in fact, my comment above explicitly states it.
Thanks for this, you've just put your finger on something that's been bugging me about a lot of the recent Mt. Gox / Bitcointgasm discussion. Actually, it raises the point further that if so many people who claim to know what they're talking about can confuse Mt. Gox's position (fraud, incompetence, insolvency, bankruptcy) with fractional reserve lending, it makes me strongly question just who the hell actually knows what they're talking about.
My understanding is that few banks were actually made insolvent, but that lots had to eat fire-sale prices due to liquidity issues. Your sketch isn't wrong, though.
> My understanding is that few banks were actually made insolvent
This was another can of worms. Mark-to-Market accounting was rescinded, partly to address this issue (http://online.wsj.com/news/articles/SB123867739560682309) but many folks thought the banks were dragging their feet on re-marking their housing assets well before that. So were they insolvent because nobody put an accurate, timely value?
Those that anticipated the situation by shorting the banks made some serious money (and then made less when the SEC banned shorting.)
In which case the banks can simply have the gov't force future taxpayers to make them solvent again. Perhaps bitcoin will one day attain "too big to fail" as well.
Reserves are the cash on hand available to pay people able to demand it. Since in practice, except during a run, only a fraction of a bank's liabilities are called in at any given time, it is safe to keep less cash on hand than the entire amount that could be withdrawn instantly.
The rest is converted to assets with equivalent value that have worse than instant liquidity, but a higher return on investment. This keeps the balance sheet honest with respect to the total value of assets and liabilities, but technically, the bank cannot honor all of its commitments in the worst case scenario. That's fractional reserve.
Mt. Gox does not even have the imaginary, illiquid assets to balance out its depositor liabilities. That makes it bankrupt, not fractional reserve.
> Mt. Gox does not even have the imaginary, illiquid assets to balance out its depositor liabilities. That makes it bankrupt, not fractional reserve.
Strictly speaking, it makes them insolvent. The fact that they are insolvent is the reason they have sought the protection offered by bankruptcy, but the two states are distinct.
While the terms are easily distinguishable to a lawyer, to others, they are practically synonymous, with insolvency nuanced towards being a temporary condition and bankruptcy nudged towards a more permanent or intractable failure.
The implication is that an insolvent person could become solvent by immediate application of better financial management, but a bankrupt person has no choice but to default on a portion of his debts or other financial obligations. Bankruptcy is the noun/adjective descibing what the court does with bankrupt people and businesses.
The ACME brand hair-splitter is the only professional-grade capillascindor you will ever need.
I recall spinning this theory on another hn thread [1]. And sure, you are correct that Mt Gox didn't make loans, didn't act exactly in all respects like bank. But the literal term "fractional reserve" is certainly suggestive of what both a standard modern bank does and what a crocked bitcoin exchange could do - only keep a small amount of money to satisfy inflows and outflows while doing something else with the rest of the money entrusted with it. If Mt. Gox did this, they clearly weren't responsible in doing it since don't have the money entrusted to them. But it pretty much seems like this lack of responsibility would be what distinguishes a failed bitcoin exchange from an ordinary banks.
Consider, Wikipedia says: "Fractional-reserve banking is the practice whereby a bank retains reserves in an amount equal to only a portion of the amount of its customers' deposits to satisfy potential demands for withdrawals. Reserves are held at the bank as currency, or as deposits reflected in the bank's accounts at the central bank. The remainder of customer-deposited funds is used to fund investments or loans that the bank makes to other customers." [2]
Which is to say, a fractional reserve system involves keeping only some money handy and hoping that the money you remove for other purposes goes on to make more money. Now, if you don't tell people you're doing this, then yes it's fraud. Secretly operating something that people don't think of as a bank, as a bank, is fraud. Lose the money you've invested and a fractional system collapses, whether you are openly operating as a bank or secretly operating as a bank. Secret banks do tend to collapse more often just 'cause they're shady. But the secret banks that make money, well you don't hear about them most of the time.
That's not how fractional reserve works; that's just fraud. With fractional reserve, a bank holds maybe 1/8 of account balances in cash and the other 7/8 in loans due to the bank. So if eight people each deposit $100,000, the bank holds $100,000 in cash and can make three mortgages for $200,000 each plus ten $10,000 car loans. Either way they hold sufficient assets to cover their deposits, just not all of those assets are fluid. (The bank earns enough interest on the mortgages and car loans that they can pay interest to their depositors and take some profit off the top.)
Exactly. And for those unfamiliar I'll add that banks face a classic problem: their loans have long terms, but their deposits can be withdrawn at any time. That is normally not a problem, but there are issues and edge cases that we handle with things like bank capital standards, deposit insurance, and swift government takeover and resolution in the case of bank failures.
For those interested in this, I strongly recommend Sheila Bair's "Bull by the Horns". She was the head of the FDIC up to and during the 2008 financial crisis, and this is her memoir. She's a fiscally conservative Republican, but one who strongly believes in the value of regulation as a way to create a sound economy so that all citizens can thrive. The book was fun to read, and it gave me a much better understanding of the forces at play and why good regulation of banks is immensely valuable to us all.
It's justified via statistical multiplexing, just like when an ISP sells a total of 1 Gbps of bandwidth but only has a 256 Mbps upstream connection.
So long as only a low percentage of demand depositors ask for their money back at any one time it's not a problem. But when everyone decides to withdraw their money at the same time you have a liquidity crisis. Liquidity problems aren't so bad anymore though -- the Fed steps in an lends all the cash you need against your long term assets. The real problem is when those assets go bad. Now you don't have a liquidity problem you have a solvency problem. The only thing the Fed can do at that point is to simply give the bank money to make up for their bad investments. Which is exactly what they did and are doing, albeit in an obfuscated manner.
Great stuff. Although I should say that there's more you can do with bad assets than just give people money. My understanding is that classic resolution is where you create a "bad bank" [1], a holding company for all the bad assets. You then fire the bank managers, put less dumb people in charge, recapitalize the banks, and use the bad bank to slowly realize the value from the assets, which are often not totally bad, just part of a cyclical slump. And you also update regulations and capital controls so as to reduce the chance of a similar mess next time.
A lot of people are critical that after the 2008 crisis very few people got fired for FUBARing the world economy. I get why the US ended up doing that; people were scared of anything that looked like more instability. But I think it was a mistake.
The Fed can't do something like that. It usually takes the involvement of a bankruptcy court, though the FDIC has some powers it can exercise independently. A full Swedish style resolution would almost have certainly required new legislation.
All that said, I agree with your underlying point that just giving insolvent banks money wasn't a great solution at all, though it was minimally sufficient to prevent bank runs at least on formal banks. There were some runs on shadow banking institutions, though in at least one case -- money markets -- the government stopped one by guaranteeing them as though they had been insured banks. That too was a mistake in my opinion.
> banks face a classic problem: their loans have long terms, but their deposits can be withdrawn at any time.
Borrow short and lend long - great work if you can get it. Of course, this practice is fundamentally unsound (it provides nasty game-theoretic incentives to participants), but it tends to work "well enough" in practice that no one really cares, especially when there is a lender of last resort who is able to print money at will.
To be fair, the lender of last resort isn't necessarily printing/creating money in the long term.
If it is simply a classic bank run, and everyone wants their money back now, the lender of last resort could take over all of the distressed bank's illiquid assets and, over time, recover some or all of the lent moneys.
In this case, what the lender of last resort is really doing is making all illiquid assets liquid.
If you think a bank run is going to happen, you want to withdraw your money ASAP, before it's all gone. Everyone else does the same - so you can get a self-fulfilling prophecy, even if everything would have been fine if everyone had stayed calm. For personal banking, this is largely mitigated by FDIC insurance, because the Fed can print enough money to cover small bank failures without anyone noticing or caring. However, if you're fucking around with collateralized debt obligations (CDOs)...2008 happens.
Banks actually pay insurance premiums to FDIC, which are held in a fund which is used to pay claims. If the fund is wiped out, practically speaking the Fed would cover it but that's not the normal method of operation.
The "Calculated Risk" blog reports on bank failure, and even in 2010 (11?) when literally hundreds of banks failed there were only a couple cases in which the FDIC had to pay anything out.
which states that "The fund had a balance of negative $7.4 billion as of Dec. 31, though that was an improvement from the $20 billion hole it was in at the end of 2009."
I was under that impression they could make loans of 72/8 of their account balances. I may very well have misunderstood, but I had thought banks could loan out a given dollar multiple times at once. No?
No. The reason there's a multiplier effect that gets talked about a lot is because the money you borrow from the bank then ends up in another bank account (of the person you were borrowing money to pay). The bank it's deposited in there can then lend out against that deposit.
But each bank loan is balanced, to the reserve ratio that bank operates at, to a corresponding deposit.
I've been saying this for a while now, including on HN.
> The reason I tend to believe this scenario is because it's completely consistent with their behavior.
Another reason to believe this is that it is something that has happened often in the past before modern banking regulations. Seriously, read history of banking in Netherlands or England, the idea:
"Oops, I just lost my customer deposits. I'm probably going to get lynched now. But hey, if I just hide this until I make the loss back from fees/reckless trading, I'll be able to pay my customers back and avoid being lynched. It's a win-win."
Is something that at least hundreds, if not thousands of bankers have had over the course of history. As far as we know, it never works.
There's no way to know how often this sort of thing has happened in the past. We only find out about the cases where a bank is not able to earn back it's loss and runs out of cash. At that point they have to admit they're insolvent.
If the bank was able to earn back the money and become solvent again, then no one outside the bank would know it happened and the bank would have every reason to cover it up to save its reputation.
> Oops, I just lost my customer deposits. I'm probably going to get lynched now. But hey, if I just hide this until I make the loss back from fees/reckless trading,
One of my favorite bits of commentary on financial matters comes from that. Some laconic British commentator was referring to Nick Leeson as "that chap who made money hand over fist for Barings Bank, until he didn't".
That's not true at all. Almost every bank in the western world did this in the past six years. They became technically insolvent due to losses in the housing bubble, but through, for example, huge piles of free money given to them by the U.S. government, were able to keep on trucking until they could earn enough money and get back into solvency. That was the entire theory behind the bailout.
Not all banks that were insolvent managed to keep on going; many failed. But many survived.
So this strategy works perfectly well if you have the ear of the Federal Reserve or another entity that can provide very large amounts of money. It works well when the amount you are insolvent by is small compared to your revenue stream. It does not work well if you are a Ponzi scheme operator and if you have stolen ALL of the money.
> The reason I tend to believe this scenario is because it's completely consistent with their behavior. I watched them very closely, and there seemed to be no rhyme or reason for their behavior. That is, unless they were missing everyone's bitcoin for some reason. Then their behavior made perfect sense.
They didn't actually need to replenish their supply of good bitcoin; they only needed to buy up all of the bad goxcoin. For example, by halting withdrawals for 2 weeks in order to drive down the price.
No. They did not become a fractional reserve, they became insolvent. A large portion of the BTC community grossly abuse the term 'fractional reserve', and it's flatly wrong.
Fractional reserves refer to liqudity, not solvency. If I run a financial institution and I owe depositors $100, but have $10 cash and am owed $100 on top of that, I'm running a fractional reserve.
If on the other hand, I'm MtGox, and I owe depositors $100, but have $10 of cash and am owed nothing on top of that, I'm insolvent. Bankrupt. And if I keep operating, I am a fraud.
Yeah, I'm still constantly surprised how so many in the BTC community seem to lack even the most basic of accounting knowledge, yet have strong opinions on how international banking and currencies should work.
> A large portion of the BTC community grossly abuse the term 'fractional reserve', and it's flatly wrong.
A large portion of the BTC community grossly abuses all economic terms because it's a libertarian anarco-capitalist circle jerk much of the time. The sane and educated are few and far between but they're there.
The important side effect of such cash loans is that once on the market, they are indistinguishable from any other cash. So the bank essentially increased amount of cash in circulation. While technically they didn't "print money", for all practical purposes, they did. And since this cash slowly propagates throughout economy, it increases the total supply which leads to raising prices. Prices raise first where this newly loaned cash is being applied. Since 2000s it is stock market. It grows in total sync with Fed's "bonds buying".
Gonna have to disagree with it being an important distinction.
In times where you can't liquidate assets at a high enough fraction of their "I deserve this much" value, then you can't meet your obligations and are thus insolvent as well. If someone's willing to buy your illiquid assets (or lend on the assumption that they're) at full "I deserve it" value, you were still insolvent -- you just got bailed out.
If you are otherwise profitable, then a long enough line of credit can return you to profitability. In that respect as well, an "insolvent" institution can become solvent thanks to this added liquidity.
For those reasons, I believe that in the interesting cases, liquidity and solvency are too deeply entangled to distinguish.
So when MtGox tries to keep the facade up long enough for trading fees to cover the shortfall, then yes, that is different from an "illiquid but solvent" bank getting a loan from the Lender of Last Resort ... but it's a different of degree, not kind (edit: fixed wording, thanks dllthomas). Both of them are trying to cover up functional insolvency with future profits they hope to operate long enough to get.
The reason I tend to believe this scenario is because it's completely consistent with their behavior. I watched them very closely, and there seemed to be no rhyme or reason for their behavior. That is, unless they were missing everyone's bitcoin for some reason. Then their behavior made perfect sense.
In that scenario, Mt. Gox would have knowingly traded non-existent Bitcoins for far, far longer than two weeks.
EDIT: I should mention that there's still no evidence whatsoever that malleability somehow led to the loss of >500,000 BTC.