Hacker News Reader

Comment by unicornmama | Hacker News Reader

> SVB has the money to pay back almost all of the depositors.

There seems to be this myth floating around that bond losses aren’t real. They are very real.

An 80 cent on the dollar (purchase price) bond is a loss of 20 cents. And it doesn’t matter if the holder holds to maturity.

Welcome to interest rates.

Edit: Fundamental fallacy here is not understanding the time value of money. Thinking of money without the time dimension is like thinking about space without time.

See https://www.investopedia.com/terms/t/timevalueofmoney.asp

Secondary fallacy here is equating value in the financial sense with gain/loss in the accounting sense.

View on HN

You don’t lose money on bonds you hold to maturity. 100% of the principal will become liquid upon maturity.

You lose money by not being invested into the better yielding options while you are waiting your bonds to mature.

Whoosh the goalposts are moving fast on this one.

What you are talking about is opportunity cost, not investment losses. If SVB could have held their bonds to maturity they would have gotten back every cent of the principal.

Yes, but you also need to look at the net present value. A loan today would need to be at below market rate interest in order to match up with the value of the bonds held to maturity.

Interest rates rising means products with a fixed rate yield are worth less today. That value isn't gotten back by waiting until maturity. The nominal value is retrieved, yes, but the money in the future is literally worth less.

SVB was very poorly run. We can see that easily now in retrospect.

The nominal value is what matters because deposits are also nominally valued.

We don't live in a magical world where deposits (aka liabilities) are exempt from inflation and the assets that back them are not.

If you owe someone $1000 you owe them one thousand dollars. Not the value or purchasing power of one thousand dollars -- literally one thousand things called dollars.

So if you take $1000 in deposits, you buy $1000 in bonds, you wait until the bonds mature, and then your depositor withdraws $1000, you will be fine no matter if that happens in one year or a hundred years, no matter what the rate of inflation is. Your assets and liabilities will cancel out.

If your depositor tries to get their $1000 back before the bonds mature, you are screwed. That is what happened to SVB. If all of SVB's bonds had been mature by last Friday it would have been fine. Pretty much every article in the financial press about this fiasco has made that point.

It is amazing to see people twist themselves up in knots about opportunity cost and inflation when this is so basic. If liquid assets equal liabilities in the literal number of dollars, you are even, you are solvent, and your depositors get their money back.

> A loan today would need to be at below market rate interest in order to match up with the value of the bonds held to maturity.

Well, it's a government loan, so they can do that if they want to.

so they would've been fine if only everyone could've waited 30 years? isn't that completely irrelevant to probably every depositor?

Of course it is irrelevant to depositors, that is why the bank failed.

However, it also proves that holding bonds to maturity is different from selling them at market rates. It demonstrates the distinction between solvency and liquidity. Every financial publication has made this point when discussing SVB in order to educate their readership about the problems of duration risk and explain how a bank with enough assets to cover liabilities can still fail.

Now, the nice thing is, the government has time to wait for the bonds to mature. So the government can take the bonds, pay off the depositors, and get the money back when the bonds mature. The government won't lose money if they do it right -- just like they didn't lose money with TARP in 2008.

It's not moving the goalposts - this is an actual, real, and extremely sneaky loss. Having bonds that will pay out say $100 in eight years time is pretty much exactly equivalent to having the reduced value of those bonds now (say $80) because if you had that reduced amount of money now you could invest it in similar bonds and receive $100 in eight years time. In fact, I think with current interest rates you could even stick it in more liquid bank accounts or short-term bills and likely still receive more interest whilst not being locked in.

They are an investment loss if you need to sell them at a loss, which seems to be the case here.

Also, at some point the difference between opportunity cost and investment loss becomes rather semantic. In a liquid market, you should be able to sell and rebuy your positions every day, which you generally don't do of course, but it does mean that the decision not to sell is similar to the decision to buy: if you wouldn't buy under these circumstances, you should sell.

With these bonds, their low interest makes them less attractive than newer bonds with higher interest, so nobody will want to buy these lower interest bonds at face value when higher interest bonds are available. They'll only buy these at a discount that would make their profit comparable to those of higher interest bonds. So the value drops, so that's a loss.

Personally I've never seen the point in buying low-interest bonds. But then I'm not a banker.

They cannot hold to maturity. Because you cannot offer your depositors 1.45% on deposits when their best alternative is 4.5-5% in risk free money market funds and treasury bills. Depositors won't just sit there and watch 6%+ inflation eat away at the real value of their deposits. The assets are correctly priced and bond losses are real.

Only a potential buyer would lose. Not SVB - for them the price and return is locked if they can wait. The problem comes when they are forced to sell (as they were).

This is why the bond price falls - an outside party will not buy the SVB bonds because they can get a better return on a different bond. The bond price falls to make them equivalent.

Better yielding options come with more risk – that's typically why they are better yielding !

Except that during the duration that 100% loses out to intlation. If it's a 10-year bond, it's worth maybe 74% of its face value.

That’s not an investment loss though. It’s really not the same thing.

If I can claim inflation as a loss I need to go redo my taxes…

It's material here because the reason Tbills are paying 5% is because inflation is roughly 6.5%. (I used a fixed inflation rate of 3% to get 74% above.) Usually they're down at like 0.05% or so, which is what SVB was holding, which is what sank them.

It’s not material.

What matters is that principal is returned in full when bonds mature, but if you can’t wait until maturity you might have to sell them for less than the principal. That is exactly what happened to SVB.

Sounds like they needed more diversity in their overall capital portfolio; based on significant risk of these long-term bond rate increases -- is this a common tactic that would be employed at other banks, but it's just that SVB had a "special" system where customers would hold more money there or something?

What's stopping my Local Bank from crashing this week?

> What's stopping my Local Bank from crashing this week?

Enough consumer confidence to prevent a bank run is the only thing that prevents a run on any bank, including the largest banks in the world.

why do you think those bonds are worth only 80% now, if you believe they should be worth 100%? is it a market inefficiency? do you think you could make risk-free 25% profit by buying them?

So do the deposits the bonds are backing. That's irrelevant.

If you need to pay someone now, the fact that in X years you have the money creates costs - whether the bonds pay 100% or not at maturity doesn't really matter, only the current MtM matters for money now as opposed to money in the future.

In addition to my comment below, the following are financially equivalent [I simplify a bit]:

1. Give 80 cents on the dollar to each depositor and say tough-luck. 2. Give $1 dollar to each depositor upon bond maturity [for sake of example let's say 10 years].

Option 1: Depositor takes the cash, buys a new bond with the same maturity [but it would have higher yield, say closer to 4%] at current market price. At maturity they have $1 dollar

Options 1 and 2 are equivalent, minus the bid-ask spreads which are very tight for treasuries. This is time value of money.

If you need to repay the bond's principal now (as SVB does), then you lose the extra interest you need to pay on the money you'll borrow today until you get the full principal back at maturity. You don't lose money if you can borrow money at the same rate as the bond coupons pay, you lose money if now the interest is higher. You can't simply treat dollars-after-X-years as equivalent to dollars-now, those are two different 'currencies' and the former is worth less.

> If you need to repay the bond's principal now (as SVB does)

Not anymore. Now the question is, how does the government make depositors whole? I doubt they will go flog the bonds on the open market or get loans at market rate like idiots. The point of this intervention is to bypass the market mechanisms that caused this and put the government as a backstop.

> extra interest you need to pay on the money you'll borrow today until you get the full principal back at maturity

Thank you, great concise explanation, lightbulb moment for me. I understood it but couldn't clearly communicate it.

So to make depositors whole today, you need to borrow money today, and the extra interest paid to borrow the money is more than the interest you get when the long-term bonds finally mature. The difference is the loss.

Which is why they had to take a loss on selling the bonds, as it's essentially the same loss. They are out money whether they sell the long-term bonds today, or if they borrow money today. Bonds are discounted appropriately by whomever buys them.

Question is, why didn't SVB do anything when they saw this coming? I've seen articles saying the board was aware of the risk issues for the past year. [1]

[1] https://www.forbes.com/sites/noahbarsky/2023/03/12/silicon-v...

Genuine question - why doesn't it matter?

If I buy $1000 worth of bonds at 2% interest rate for 10 years, my expected return is 1000 * .02 * 10 or $200, making the bond worth at maturity $1200. This bond is worth $1000 today and will return $200. If the bond price falls to $.80 on the dollar or $800 and I am forced to sell today to make my depositors whole, now there is a realized loss - $200 from the original price and $200 from the eventual returns.

If I can wait I have $400 more. If I can't and have to sell then I lose $400.

You are absolutely right.

The distinction between insolvency and illiquidity is a red herring. People who have to sell their house in foreclosure will be discovering the exact thing you are describing here.

What's happening here is SVB is big enough that the authorities go "aaaaah, wait a second, this could blow up a bunch of other businesses and we wouldn't want that". And so they are taking the illiquidity interpretation and helping out the depositors, but at least they are not helping the shareholders.

One thing that may be a bit different in the case of a bank is that the general public does not know or consider deposits to be a loan to a bank. Which is what it is, but people don't think of it this way, and the aren't encouraged to either thanks to FDIC and other state level guarantees. If we change that by letting the depositors lose money, there's going to be chaos.

Here’s the thing: when illiquidity leads to temporary insolvency, it’s easy to chalk it up to bad luck. SBF would probably agree. But an Austrian economist would probably argue that no, actually what is happening is that market forces have determined that you made bad long-term investments and they ought to be liquidated sooner rather than later so the economy can shift that capital into more productive uses.

You can't just add up future value and current value of things. The future value is in future dollars, they are different from current dollars. You owe depositors their current dollars.

Yes you owe the depositor in current dollars, and that's why the bank failed. But I was responding to the parent about it not mattering if the bond is held to maturity - it does matter.

The problem SVB had is that there was no market buyer for their bonds at a price they needed today. They deserve to fail for that but that's not the part of the discussion I am responding to.

What I am responding to is the idea that there is a myth about the value of the bond. Here is what might happen, in a very simplified way:

- Depositors need their cash today - SVB can't sell their assets to meet this need, and so the bank is fails and is dissolved (already happened). Let's make this simple and say SVB owes the depositor $1000, can sell for bonds for $800 today. If they can have wait the bonds will return $1200 later. - The FDIC steps in with all their capital. They say ok - depositor here is your $1000 today and you are now whole. But we will not sell the SVB bond today to cover that $1000, instead we will hold the bond and wait for it to mature at $1200. Thus the depositor is whole, and over the long term no money is lost.

No regular market participant step in to provide the $1000 because they can get a better return on their money in other ways. But the government can do this because their goal is not maximizing return on capital, but instead stabilizing the system.

The government needs to get its money from somewhere. If it spends the taxpayers' money, that money cannot be spent on other things. So instead of doing things that are useful to society, like maintaining roads, the money is just sitting there until the bond matures. If it creates money out of thin air, the effect is the same, except that now every market participant pays (in the form of increased inflation). So in either case, the losses are socialised.

Of course, you can still argue that stabilising the system is worth it.

$1000 worth of bonds at 2% simple interest rate returns $1200.

The US 10 year treasury interest rate is 3.7% right now. That means you can get $1200 in 10 years with $835 today at 3.7% compounding; 1200 / 1.037^10 = 834.44 and change.

So your $1000 worth of bonds is actually only worth about $835, at best, because that's the market price for a (close as possible to) risk-free investment which matches the return at maturity.

Inflation is the flip-side of this. You can reasonably expect $1200 in 10 years to be worth about what $835 is today. It might be less, it might be more, but it's an estimation with money behind it.

> risk-free investment

That's what I don't get. Why are bonds considered risk free if their value can drop when interest rates go up? Sure, they may be worth $1200 in 10 years, but they're only worth $835 now, when they were worth $1000 yesterday.

Risk may be lower than buying shares in a company at risk of bankruptcy, but it's hardly risk free. These things can go up and down just like normal share prices.

There’s no default risk, that’s all. “Risk-free” only refers to the risk of default. Every (fixed-rate) debt instrument has unavoidable interest rate risk. And the floating rate ones are just transmuting it into default risk.

You are correct, these are two different kind of risks. The "risk-free" rate refers to counterparty risk (which should be zero when the counterparty has the money printer, or can be bailed out by the money printer.)

In your example, the market's saying that $1200 in 10 years won't have the purchasing power of $1200. And the future purchasing power would be closer to $800 than to $1000. Consider an alternative world where inflation is zero but you paid $1200 for a $1000-par bond.

The important concept here is the time value of money.

This is not how investment losses or returns are calculated for accounting purposes, which is what matters when we talk about the solvency of a bank.

The price doesn't fall because the future purchasing power falls, the price falls because there are better alternatives in the market. No one will pay $1000 for a 2% return when they can pay $1000 for a 5% return. Market participants will always maximize their return. Bond prices adjust to be competitive or equivalent.

But the govt. doesn't have this problem. They don't care about maximizing return - they care about containing contagion. So they can pay $1000 for a 2% return and not still not lose money over the long term.

Inflation strongly determines interest rates. You are getting 5% return because inflation expectations are high which caused the Fed to raise interest rates. When you see 5% risk free, you assume high inflation. The price falls because of the risk free rate, which is caused by high inflation, which causes drop in purchasing power.

Reply on news.ycombinator.com