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

Mortgage-backed securities III

More on mortgage-backed securities. Created by Sal Khan.

Want to join the conversation?

  • male robot johnny style avatar for user Chris Archer
    So does this mortgage backed securities methodology continue today for non-traditional loans? Or has it gone back to normal? What about in China?
    (24 votes)
    Default Khan Academy avatar avatar for user
  • blobby green style avatar for user alex.wickens
    At Khan seems to think that the net effect of 20% of people defaulting with a 50% recovery rate will lead to only a 10% decrease in returns. This is confusing because I did a crude calculation (assuming all defaults happen at end of year and defaulters pay interest until they default), and I got that at the end of the first year you would have 100M from house sales, 100M from interest and 800M owed to you.

    In other words, it seems that the return would at best 0%. Am I mistaken?
    (6 votes)
    Default Khan Academy avatar avatar for user
    • blobby green style avatar for user BradNayles
      I think the math is actually a little more complicated and the new rate of return will actually be less than 9% but not 0%.

      If all $1B of loans are being repaid, then the return is $100M per year and you'll get $100 back (or 10%) on your $1000 share.

      If 20% of the loans default and there is zero recovery, the stock just lost $200M in assets and only $800M is now being repaid. Thus the return is $80 (8%) on your $1000 share.

      If 20% of the loans default and there is 50% recovery, the asset lost only $100M; however, only $800M is now being repaid just like in the previous example. The $100M recovered does not generate any income because those debts were retired when the properties were sold. Until this $100M cash-on-hand is re-invested the return is $80 (8%) on your $1000 share.
      (9 votes)
  • blobby green style avatar for user Vinay Vasyani
    Didnt you say in the last video that MBS would be bought at 1100 per share and so the Investment bank makes the bandaid...now you said its sold for 1000 per share and investors get interest...whats the clarification...investors buy for getting the interest, if so then Investment bank wont make any money
    (4 votes)
    Default Khan Academy avatar avatar for user
    • mr pants teal style avatar for user vwerre
      Keep in mind it's a multiple year stream of payments (30 yr mortgage), the investment bank marks up the MBS by some amount (from 1000 to 1100 in this example) then sells it. Therefore, the purchaser of the MBS has paid a slight premium to own the shares, and will receive slightly less than the 10% yield the bank was making before the mark up and sale. The buyer paid 1100 for the MBS because the $100 per yr payment until the MBS matures is worth it to him.
      (5 votes)
  • blobby green style avatar for user fabio.gaucho
    I like this series of videos. But why don't we explore these banking operations further by making a series on how things like insurance, stocks, government bonds originated and developed? The picture we are getting is of a system that already became very complex. Why not show historically how new operations were developed? I am sure you could make it very interesting.
    (4 votes)
    Default Khan Academy avatar avatar for user
  • blobby green style avatar for user Arthur Segovia
    Who initiates the securitization of the grouped loans? Does the investment bank tell the commercial bank that they will buy the loans?
    (1 vote)
    Default Khan Academy avatar avatar for user
    • mr pink red style avatar for user Nathan Muller
      It can be seen as a cycle. Banks always have a certain demand for making loans; namely, the total number of loans that could be profitable in a given area for a local bank. When they could sell these loans and offload the risk to someone else (in this case an investor of MBSs), their focus shifted to sell as many loans as the investor would buy. If investment banks stopped buying mortgages, then the bank wouldn't be able to offload any increased risk they had taken, and would revert back to making the number of loans they 'normally' would have.
      (6 votes)
  • blobby green style avatar for user Benson Igarabuza
    At , Khan mentions that a 20% default and a 50% recovery on the mortgage yields an average 10% of loans being worthless. I don't get the math behind it.
    (1 vote)
    Default Khan Academy avatar avatar for user
    • blobby green style avatar for user alex.wickens
      The maths behind it is just multiplication, if 20% of the loan loses 50% percent of its value, then 0.2*0.5 = 0.1 of the value of the total loan is lost.

      I'm afraid the answer given by alexander.bendavid contains another flaw, the SPE is not assumed to able to recover half the value of the house, but rather half of the value of the initial loan in cases of default.
      (8 votes)
  • blobby green style avatar for user wngasinur
    do actually borrower know that if they are taking mortgage with S.P.E or normally bank wont tell the borrower ?
    (0 votes)
    Default Khan Academy avatar avatar for user
    • mr pink red style avatar for user Nathan Muller
      Banks will always reserve the right to sell their right to the loan (interest payments plus principal repayment) they originate for you. As a borrower, it adds no additional risk if your loan is packaged and sold, it just allows an investor to 'invest' in you ability to repay your debt. There are other tangents to go off of, but that is the simple answer
      (5 votes)
  • blobby green style avatar for user Chris Tao
    the money don't add up. home owners borrowed the $1B from the investment bank, whom then creates and sold shares of special entities for $1.1B. From the flow of money, the investment banks already got return on investment. When home owner foreclose on the loan, investment banks takes ownership of the house; which in essence a FREE house. So, why we need to bailout the banks during the housing crisis? I don't notice the investments banks losing any money.

    NOTE: the stock holders don't own the houses.
    (1 vote)
    Default Khan Academy avatar avatar for user
    • male robot hal style avatar for user Andrew M
      The banks did lose a lot of money. If you lend $1 million and then you get back a house that you can only sell for $500,000, you lost money. That doesn't mean they should or shouldn't be bailed out, of course.
      (3 votes)
  • mr pants pink style avatar for user KarlKarlJohn
    Is the recovery percentage always the same regardless of when the borrower defaults on the loan? For example, if I default on a $1 million home loan after having made payments for many years, taking the remaining principal down to say $600,000 is the recovery percentage still 50% if the bank sells my home for $500,000. Or would it be 90%, the $400,000 I had already paid off plus the $500,000 the bank got for selling my home? Or something else entirely?
    (2 votes)
    Default Khan Academy avatar avatar for user
    • male robot hal style avatar for user Andrew M
      The recovery percentage in your example would be 500/600.

      Normally you do not see recovery percentages like that because if the bank can sell it for $500, a more patient seller might be able to sell it for more, and therefore the owner is unlikely to default rather than just sell the property. But it could happen.
      (2 votes)
  • mr pants teal style avatar for user Joe St
    Do we have an option to short on these MBS investment ?
    (2 votes)
    Default Khan Academy avatar avatar for user

Video transcript

Welcome back to my series of presentations on mortgage-backed securities. So let's review what we've already gone over. So I've already drawn here-- I actually prepared ahead of time-- so I've already drawn here kind of what we've already talked about. So we start with borrowers who need to buy houses. Each of them borrowed $1 million. Actually let me write that down. Let me change the color of my pen. Where'd my pen go? OK. So each of these people borrowed $1 million. OK. Each of them borrowed $1 million and there were 1,000 of them, right? So $1 million times 1,000. That's $1 billion that they needed. And they said that they would pay 10% a year on that money that they borrowed. So that's 10% for each of them is $100,000 and then as we said, there's 1,000 borrowers. So they're going to put in $100 million, right? 100,000 times 1,000 is 100 million. So just to simplify. Keep it in your mind. $1 million goes to a bunch of borrowers, goes to 1,000 borrowers, to be specific. And then each year, those borrowers are going to give the special purpose entity-- this is just a corporation designed to kind of structure these mortgage-backed securities-- they're going to give 10% of the billion, or $100 million back into this. And then we said, OK, well where does that money for this special purpose entity, or for this corporation, come from? Well it comes from the investors in the actual mortgage-backed securities. And just to be clear, so the asset within this entity are the loans. The loans are the main asset that's inside of the special purpose entity. And the loans are just the right on these 10% payments. And so money came from when the owners of each of these mortgage-backed securities-- each, let's say, paid $1,000 for the mortgage-backed securities. And in return, they're going to get 10% on their money. So each security cost $1,000. And then they're going to they're going to get $100 back per month. And we said there are a million of these securities, so $1,000 times 1 million, that's where the $1 billion comes from. My thing's been acting up. That's where the $1 billion comes from. And that essentially is lent to the borrowers. And these guys will get 10%. Now one thing I want you to keep in mind is, they get 10% only if every one of these borrowers pays their loans, never defaults, never pre-pays. Pre-paying a mortgage is just saying, I sold the house. I don't need the mortgage anymore, so I just pay it off. So it's only 10%, indefinitely, if all of the borrowers pay all the money and never default or anything like that. So this 10% is kind of in an ideal world. Well everyone knows that it's not going to be exactly 10%. Some percentage of these borrowers are going to default on their mortgage. Some of them are going to pay ahead of time. And actually that's what the buyer of the mortgage-backed security should try to figure out. And all sorts of buyers are going to have all sorts of different assumptions. And this is what you probably read some articles about, these hedge funds with these computer models to value their mortgage-backed securities. And that's what those computer models do. They try to look at historical data and figure out, OK, for a given population pool in a given part of the country, what percentage of them are able to pay off their mortgage? What percentage of them default on their mortgage? And when they default, what is kind of the recovery? Say they default on a $1 million mortgage, and then the special purpose entity would get control of that house. And then if that house is sold for $500,000 because the property value went down, then the recovery would be 50%. So that's all of the things that someone needs to factor in when they figure out what will be the real return. 10% is if everyone pays. So let's make some very simple assumptions for ourselves. Let's say we are thinking about investing in a mortgage-backed security and we want to gauge for ourselves what we think the return is going to be. Well let's say we know that this pool of borrowers that-- my pen keeps not working-- that 20% will default. We're not going to worry about pre-payment rates and all things like that. Let's say 20% are going to default. Of these 1,000 borrowers, 200 of them are just going to lose their job or whatever. They can't afford a mortgage anymore. And of those 20% that default, we have a 50% recovery. So that means borrower X defaulted on his loan. And then when we go and get the property-- because the loan was secured by the property-- when we auction off the property, we only get $500,000 for it. So we get a 50% recovery. 50% of the original value of the loan. So if 20% default and then there's a 50% recovery, then on average you're going to get 10% of the loan is worthless. And I'm going to make some kind of handwaving assumptions here. But you can assume statistically, and since this is a large number of borrowers-- it's 1,000, right? If there's only one borrower it would be hard to kind of gauge when he defaults, if he defaults at all. We would just know that there is a 20% chance. But when there's a large number of borrowers, you can kind of do the math and say, OK, on average 200 of these guys are going to default, and instead of actually getting 10%, since 10% of the loans are going to be worthless, I'm going to get 10% less than this 10%. So I'm going to get 9%. So this is based on the model that we just constructed, right? This is the model that we constructed. This is a much simpler model than what most people use. But based on the model that we just constructed, I think the real return we're going to get on this mortgage-backed security is 9%. If there was another investor who assumed a 50% default rate, but with a higher recovery, he or she would have a different kind of expected return from this security. So why is this even useful? Well think about it. Before, in the case we did in the first video, when someone just borrows from the bank, the bank has very specific lending requirements. They have their own model. So there's a whole class of borrowers that they might have not been able to service. Right? There might be people with really good credit scores, really good incomes, who don't have a down payment. And if they don't meet what the bank's requirements are, they would never get a loan. But there are probably some investors out there that would say, you know what? For the right interest rate and for the right assumptions in my model, I'm willing to give anybody a loan, as long as I'm compensated for it enough. And this is what this mortgage-backed security market allows. It allows-- let's say this group of borrowers-- let's say this pool of borrowers right here actually didn't-- This pool of borrowers actually aren't the traditional-- they don't have 25% down and they don't have kind of the traditional requirements to get a normal mortgage-- but if I pool a bunch of people who don't have those traditional requirements, but they're good in other ways-- they have a high income or high credit score-- I can go through this alternate mechanism to find investors that are willing to loan them money. So essentially, from the borrower's point of view, it allows more access to loan funding that they would have otherwise not been able to. And from an investor point of view, it allows another place for me to invest in. Maybe I feel that the computer models that I have are really good at predicting things like default rates, and recovery rates, and what a loan is worth. And I feel that I can, in some ways, be a better loan officer than the banks. And this would be an attractive place for me to invest in. It also might just have a risk reward characteristic that doesn't exist in the market already, and it allows you to diversify into one other asset class. So that's the value that it has across the entire spectrum. Now in the next presentation I'm going to show how you can, I guess, further complicate this even more, so that you can open up the investment to even a larger group of investors. Because you can think about it right now, there's probably some people who say, OK, I already said, some people will do these models and try to make their own assumptions and say, OK, this is going to give me 9% a year. But there's a whole bunch of people who are going to say this is just too complicated for me. This seems risky. I don't have any fancy models. I only like to invest in things where I know I get my money. Very highly rated debt is where I'm going to invest my money. And there's another group of people who say, OK, 9%, that's nice and everything, but I'm a hotshot, I'm a gambler. 9% isn't the type of returns I want. I want to take more risk and more return. And so there should be something, maybe, for those people as well. So that's what we're going to show you in the presentation on collateralized debt obligations. See you soon.