If you're seeing this message, it means we're having trouble loading external resources on our website.

If you're behind a web filter, please make sure that the domains *.kastatic.org and *.kasandbox.org are unblocked.

Main content

FRB commentary 2: Deposit insurance

More on the weaknesses of fractional reserve banking. The FDIC and deposit insurance and its side effects. Created by Sal Khan.

Want to join the conversation?

  • leaf green style avatar for user Samuel Maniquis
    If the notes (money) does not represent the gold anymore, what does the Fed still "owe" to us in that IOU note? Hence, can we say otherwise then about money? I just thought if when I got through the video somewhere at ...
    (8 votes)
    Default Khan Academy avatar avatar for user
    • male robot hal style avatar for user Greg L
      The IOU's Sal is talking about do not pertain to the dollar bills in your wallet. The IOU's mentioned are from loans between the Fed and the banks it does business with.
      The Fed owes you nothing for a Federal Reserve Note. Federal Reserve Notes are a fiet currency, the reserve notes are not an IOU, they have no redemption value with the reserve, they are only worth the value of the goods or services you can trade them for.
      (6 votes)
  • blobby green style avatar for user sara.belal2003
    Sal forgot about the bank where he wrote "100" what about that bank?Is bank 4 bank100????I don't understand!
    (I'm only a kid!). (with an iPad.............) @@
    (3 votes)
    Default Khan Academy avatar avatar for user
    • leaf green style avatar for user MonicaPe
      Sara, you are a very smart kid! That bank is included in his example. Sal got a little ahead of himself. He is just trying to demonstrate the "domino effect" of how one bank can fail after another. Hint: Use the closed captioning even if you are not hearing impaired as that helps reinforce what Sal is saying and also helps with spelling. Keep up the great work, Sara!
      (15 votes)
  • leaf green style avatar for user daniel
    when the FDIC gets fees for insuring banks, where does that money go? The treasuery? Back into the federal reserve?
    (4 votes)
    Default Khan Academy avatar avatar for user
  • leaf green style avatar for user Palfi Andras
    So when times are bad we would have to pay(trough taxes) for our own money,right?
    (5 votes)
    Default Khan Academy avatar avatar for user
  • blobby green style avatar for user hnminh24
    what is the difference between federal reserve note and federal note outstanding? is note outstanding the same as cash? if so, why at Sal said that Fed lend out money to member banks by giving them FRN? because if I want to withdraw my money, I would want cash and not US Treasuries notes..
    (2 votes)
    Default Khan Academy avatar avatar for user
    • blobby green style avatar for user john
      Cash = Federal Reserve Notes. Paper currency in the US used to be issued by congress but now it is issued by the Fed (a private corporation). These 'dollars' are actually called Federal Reserve Notes because they are notes from the Federal Reserve. You would not withdraw treasury notes as those are an investment like T-Bills or T-Bonds (the difference between the three is only the duration to maturity).
      (4 votes)
  • blobby green style avatar for user jma4
    Wouldn't the banking system be more sound without FDIC insurance? Without deposit insurance people would actually care and take time to look into where they are throwing their money. In addition, banks would be forced to compete for deposits and draw people in with an interest rate appropriate to the level of risk desired by the depositor.

    Why not allow the private sector to supply depositors with deposit insurance if they want it?
    (2 votes)
    Default Khan Academy avatar avatar for user
  • blobby green style avatar for user jy.weissberg
    Why do we assume that FDIC insurance is underpriced? If financial insurance is difficult to price due to correlations, isn't it just as likely that the FDIC insurance is overpriced?
    (2 votes)
    Default Khan Academy avatar avatar for user
  • male robot hal style avatar for user Michael
    I see that the FDIC undercharges the big banks for the insurance, and the safety net it offers the banks encourages them to take on additional risks to outcompete other banks. In this type of environment, might it be a good idea to adopt a fluctuating premium, where the costs of FDIC insurance go up for a bank that is increasing risks, and down for a bank that is decreasing risks?
    (2 votes)
    Default Khan Academy avatar avatar for user
    • leaf green style avatar for user Ryan
      In a perfect world your example seems pretty logical. Unfortunately there are a lot of problems with such a system.

      First of all, major problems are never evident until after the fact. The last recession is a perfect example of that. So coming up with a way to judge such risks may be impossible. Also, the government is dreadfully outmatched in such circumstances. It would take a team of 25+ people working full time just to adequately stay on top of one of the six largest banks. Those kinds of resources just aren't going to happen. Even the for profit ratings agencies can't compete. I remember hearing that leading up to the last recession the ratings agencies each only had a team of about 5 people covering the entire banking sector.

      Also, the line of who qualifies for FDIC insurance is getting really blurry. Banks that don't carry much in the way of deposits (Morgan Stanley, Goldman Sachs) are now able to tap into FDIC insurance, when previously they wouldn't have qualified.
      (2 votes)
  • blobby green style avatar for user josh.nrb
    I understand that if people deposit their own money that they should be able to expect to withdraw it whenever they want, but that's offset by actual reserves. It seems to me that for a bank run to happen most of the people coming to withdraw their gold are the people with loans and it was never really actually their money to begin with. If the problem is the money the banks lends out itself then why doens't it just make it a condition of the loan that you can't withdraw all of it at once?
    (2 votes)
    Default Khan Academy avatar avatar for user
  • leaf green style avatar for user Amandeep Singh
    When Sal talks about reserves of the bank being depleted, in modern times what are those reserves? Clearly gold is out, so is Sal talking about Cash? What constitutes the reserves of modern banks?
    (1 vote)
    Default Khan Academy avatar avatar for user
    • male robot hal style avatar for user Andrew M
      Reserves are just any deposits that the bank has taken in that it hasn't lent out. Those funds may be placed on deposit at the fed, or invested in treasury securities, or held in cash in the vault, or invested in some other way, depending on what the bank's overall reserve ratio is and what laws govern it.
      (2 votes)

Video transcript

In the last video, I touched on the core weakness of fractional reserve banking and that's the idea of the bank run-- where if you have a bunch of banks and they've all lent out 90% of their reserves-- let's say that here's all my banks-- Bank One, Bank Two, Bank Three-- and you could have a ton of banks. I mean, you could imagine you go all the way to Bank 100. The problem with a bank run is if just even one bank all of a sudden isn't able to give its depositors its money-- so it is illiquid-- and we'll reserve judgment as to whether it is truly insolvent, but it really just doesn't have the money to give it to depositors, then all of the depositors of all of the other banks get scared and run over to the banks and ask for their money back. And by definition, in a fractional reserve banking system, these guys only have 10% of their deposits as reserves, 10% reserves. So if more than 10% of the people ask for their money, they're all going to be illiquid. Then you're going to have just an all-out panic. You're going to have an all-out panic of the financial system. As soon as any of these guys' loans come due-- so let me draw one of their balance sheets-- let's assume that they were actually good investors, which is a big assumption, but let's say that their balance sheets look like this. Here's all of the deposits that they owe to people-- to me and you. That's their liabilities. They have 10% of their assets as actual currency reserves. And then the rest is loans out to businesses and whatever else. And that's an asset because businesses owe them something. And what's going to happens-- as soon as there's a panic, everyone wants their money out. The reserves are depleted very quickly. And then these guys sit there waiting for their money. Essentially it shuts down the bank. As soon as these loans mature-- maybe this guy owed his money in a year. Then they're going to say, just give me the money back. I'm not going to renew the loan even though you have a good business. And so everyone's going to want their money back and the whole financial system's going to break down and credit is going to freeze. And this is a huge problem. Remember, this could be just because by one bad actor who couldn't manage their liquidity properly. Even worse, it could be caused just by one panicky withdrawer. There's doesn't even have to be a bad actor. It could just be some fear that enters the system-- and even for a completely legitimate bank-- and I'll use legitimate in quotations because we're assuming legitimate-- because some would debate whether fractional reserve banking is legitimate. But even if this is a completely up-and-up bank, if for whatever reason more than 10% of their deposits get scared, then all of a sudden they'll be illiquid and you'll be in this all-out panic condition. Now I talked in the last video, some of the things that have been engineered to fix this problem. So these are the fixes. The first one is to have a banker of last resort, a lender of last resort. And that's our central bank-- or the Federal reserve bank. And they'll never run out of reserves, because these reserves can be borrowed from them and all these are, are IOUs from the Federal reserve. So what the Federal reserve does-- and I've talked about this in the past-- let's say that their balance sheet currently looks like this. Let's say that this is their current liabilities. These are their current assets. And their current liabilities are essentially going to be IOUs. These are going to be their reserve notes, but let's say someone else comes and says, hey, federal reserve, lend me some money. So what they're going to do is that they're going to print some-- so these are Federal reserve notes. And then they're going to just create an offsetting liability and they have no reserve limits. Let me make it a new color just to ease the monotony. So they'll have this corresponding notes outstanding liability and they're going to give those notes to you. They're going to give those notes to the bank that's in a desperate situation if it couldn't borrow from any other bank. And so the Federal reserve balance sheet will now look like this. This is what it looked like before. It has these new notes outstanding and now this will say-- instead of federal reserve notes in its vaults, it'll now have this little thing on its books called loan to the bank, loan to this guy right here, which is an asset because he owes him something. And there's no limit to which the Federal reserve could do this. Obviously there might be some kind of implicit limit. If they keep doing this arbitrarily, people might not accept these Federal reserve notes as actually carrying value. So that's probably the main limiting concern. And I'll probably talk more in future videos on other things that are kind of keeping the Federal reserve from just doing this willy-nilly, but the general idea is that they never run out. So this is one idea-- lender of last resort, but not even the Federal reserve wants to keep doing this because what if this guy is essentially insolvent? If these loans are actually going bad, then there's no reason that even the central bank should actually be lending to this guy. They shouldn't be lending to this guy. So let's say the Federal reserve looks at this guy's books and say, you know what? These loans are really worthless. I'm not going to lend more money to you. What I'm going to do is take you into receivership. And I haven't even talked about what that is. That's the second fix. The second fix is to have this notion of insurance, of deposit insurance. So if a bank just has a liquidity problem-- it's a good bank-- it just ran out of reserves and no other bank is willing to lend to it-- which you might debate whether it's even a good bank if the other banks aren't willing to lend to it-- then it can go to the discount window at the Federal reserve and borrow some reserve notes. Now, in the situation where even that's not good enough and the bank is essentially out of business, it had been paying the FDIC a little bit of money-- the Federal Deposit Insurance Corporation-- a little bit of money-- I mean, probably a lot of money by an individual's point of view, but a little bit of money from a bank's point of view-- and it's a fraction of a percent-- a little bit of money every year so that it could tell its depositors, look, even in the situation that our bank goes bust, the Federal reserve is going to pay you directly. These are FDIC bank accounts. The Federal reserve will make you whole completely, regardless of what happens to the banks. Now this, to me, is a big fix. I mean, I'll tell you right now, all of my savings right now are in Federal deposit insured-- FDIC insured-- bank accounts. This might sound great to the depositor, but what's the side effect on the banking system? Let me draw all of my banks again. So I have Bank One, I have Bank Two, I have Bank Three, I have Bank Four. Now, they are all FDIC insured. They pay some small fraction of a percentage every year so that they can tell their depositors, look, if we ever go bust, if we ever blow up for whatever reason, your deposits are 100% safe. And what's going to happen? This will solve the bank run problem, right? Let's say for whatever reason this guy runs out of reserves or he goes bankrupt-- let's say that this guy goes bankrupt. In the old situation, before the insurance, everyone else is going to get scared and you're going to have a bank run. They're going to ask for their money and there's just not going to be enough reserves because they only keep a fraction of the reserves. That's what fractional reserve banking is, but now everyone's going to feel comfortable, like, you know what? Everything is backed by the Federal reserve, which can print as many notes is it wants so I'm just going to sit pretty. So this does solve the bank run problem. And so you might say, this doesn't make it an unstable equilibrium any more-- which is kind of true, but what does this do? All of these guys are now FDIC insured. I'm equally willing to give my money to any of these guys. So I'm going to essentially going to give my money to the guy who gives me the highest interest. And I've actually done that with my personal accounts. I've gone and I've seen which bank provides the highest interest and I've ignored how they're investing their money-- I mean, otherwise I'd be pretty suspicious about a bank offering more interest because they're probably doing something risky, but now I don't care because I know it's FDIC insured and even if that bank blows up, I'm going to get all my money back. So I might as well give my deposit to the bank with the highest interest. Now based on what you saw in that last video, which is the bank that's going to be giving the highest interest? What's going to be the bank that's taking on the most risk? So once again, when you give this relatively cheap insurance to all of these banks and all of these banks can attract money much easier by paying this insurance-- if they didn't have this insurance, they would have to pay much more. Let's say this insurance is 0.1% per year. If they didn't have this insurance, they would have to pay much more than 0.1% per year to the depositor in interest to make up for the risk that they're taking on, that they're not being backed by the government. So it's a really good deal for the depositor. Let's put it this way. When you give a deposit to a bank, you're lending money to the bank. So your checking deposit rate-- so let me write this down-- your checking deposit interest rate is the same thing as the bank's borrowing cost, right? When I go to some bank and they're paying me 2% a year, that's the bank's borrowing cost. They're the same thing. Now if the bank didn't have FDIC insurance, it wouldn't have to pay 0.1% a year, but it would have to give me more interest. It would have to pay me 3% or 4%, probably more, because I'd say this is a risky bank-- or it's more risky. Any bank would be more risky than someone who can print out the notes. So it would have to pay more, but now it can just pay 0.1% a year or whatever the Federal Deposit Insurance Corporation pays it, but it's going to lower to FDIC, but the main side effect of this is it's going to lower what it has to pay the depositors because the depositors say, this is a lot safer than what I thought. So it's going to lower borrowing costs by much more than 0.1%. So it's a really good deal for the bank and it's an especially good deal for the ultra-risky banks who might have to pay a little bit more, but because they can now do super duper risky things. You say, hey, but this is insurance. The Federal Deposit Insurance Corporation isn't stupid. It should just charge more for riskier banks. But I guess the rebuttal to that is that's easier said than done because when times are good, like in the '20s or during the housing bubble, you get hidden risk. So you don't know who's taking on the risk. It's usually the person who's getting the highest returns, but if they're always good for it, you don't see it and that person just looks like a genius. And then when times are bad, that guy blows up, but guess what? He's not on the line for it. It's the Federal Deposit Insurance Corporation, who all along was probably undercharging for that insurance. I want to make one last point about this. It's called insurance. It's the Federal Deposit Insurance Corporation. And I want to point out the difference between financial insurance-- and you could watch my video on credit default swaps, which is another form of financial insurance-- between financial insurance and what we normally associate with insurance. If I have car insurance-- let's say I'm a car insurance person and I agree-- let's say that they are Driver One-- let's say I have a bunch of drivers-- so Driver One, Driver Two, Driver Three, Driver Four. And I know that in any given year, they each have a 10% chance of an accident. That's a lot higher than reality, but we'll make the numbers easy. And if there is an accident, it will cost me $10,000. So I could go to Driver One and say, look, there's a 10% chance you're going to have an accident. It would cost $10,000 if you do. So I'm going to charge you-- the fair price would be $1,000. I'll charge you $1,100. I'll charge $1,100 just so I can make a little bit of a profit, right? What you would charge is $1,000 if you wanted to break even, but I'll charge $1,100-- and I might do this more than just four drivers. I might do it with 4,000 drivers. So on average in any given year-- let's say I have one million drivers-- in any given year, how much am I going to bring in? I'm going to bring in a million times $1,100, which is-- I'm going to get $1.1 billion in input, or in revenue, from people paying their premiums. And how much am I going to have to pay out? Well, roughly 10% if you believe these statistics are going to have an accident. So that means that 100,000 are going to have an accident. Each of those 100,000 accidents are going to cost me $10,000. $10,000 times $100,000. I'm going to have to pay out $1 billion for the accidents. So I will net every year-- if my actuaries did their job, I will net $100 million every year. And this works out pretty good assuming these statistics are right and people are willing to pay the premium because if you have enough drivers-- like a million drivers-- the statistics really hold. In any given year, the chance of 20% of the drivers having an accident is very, very, very-- almost infinitesimally small. It's almost zero. And one driver having an accident doesn't make it more or less likely that another driver will have an accident when you're talking about on the order of a million drivers. And that's true of health insurance and everything else. It's just purely probabilistic and if you have enough people insured, the statistics can actually hold up. Now with financial insurance, you have a different situation. You might say, look, in any given year, on average, the FDIC might say, look, only-- I don't know-- one out of out of 1,000 banks fail. So there's a 0.1% chance of failure. Of course it'll capture back some money because it'll have some assets on average. But the reality is that when times go bad, the bank failures tend to be correlated because they're all interlinked. And I've done a lot-- one might be lending to some business that's dependent on some business that this guy's lending to. The whole financial system is all linked up. So the failures aren't-- they all correlate with each other. They're not independent events. So if one bank fails, it's much more likely that other banks are going to fail all at the same time and we're experiencing that right now. So you can't really follow the insurance model because the statistics really don't apply. You can do this insurance model when each of the events that you're insuring are uncorrelated, but when all the events that you're insuring are likely to all happen at the same time, what you're going to have is-- when times are good, you're going to see a much lower kind of loss rate and then when times are bad, all of your guys are going to blow up and you're essentially on the hook for all these people. We see that right now and the FDIC is actually-- it's going to have to go back to Congress, I'm 100% sure, and ask for more money and it's we the taxpayers that are going to be on the line for this type of thing.