Why are crypto bankers so bad at banking? For the last several weeks we’ve read as multiple crypto bankers, starting with 3AC, then Celsius, then BlockFi, and now potentially Voyager (which recently suspended withdrawals) have gotten themselves into trouble. It’s a bit of a problem if crypto finance is supposed to replace traditional finance as a way to provide credit and run our world’s economy. This isn’t exactly what we all had in mind.

The sad thing is that the errors being made are not new, novel or exciting (you’d hope that in crypto we’d at least be discovering new mistakes to make). They’re the same ones we have known about for decades (sometimes centuries) - bank runs, lax credit underwriting (along with its evil twin concentration risk), maturity mismatches, counterparty risk, contagion. Look at any of the recent bankruptcies, bailouts, freezes and the like, and you will find one or more of these concepts at work. And behind the concepts a CEO whose business plan was basically, up only.

There’s this idea of Chesterton’s Fence. It’s a worthwhile idea. The idea is this: you buy a large piece of rural property, many acres of rolling hills and lush valleys. One day, as you are walking around your newly acquired land you find a fence extending for miles in both directions across a beautiful valley. And you say to yourself, this fence is a bit of an eyesore, and really gets in the way of a nice walk. Should I get rid of it? And so this is the thought experiment - what are the arguments for and against removing the fence?

The arguments for removing it we’ve already made. It’s an eyesore, it gets in the way of a walk. What are the arguments against? Well, there aren’t really any, are there? You just bought the land, you don’t really know why the fence is there, you certainly don’t have a reason you can point to, do you? This is where Chesterton would step in and say there actually is a reason against tearing down the fence. The reason is that the prior owner was presumably a rational person, and when they put the fence up they probably had a reason for it. So, before removing it you should try to explore that past owner’s reasons, try to discover why they would want a fence there, and only after you have done that should you consider removing the fence.

I’m sad to say that in the crypto world we have often taken the approach of tearing down the fence immediately, and then being surprised by the often very good reasons that it was there to begin with.

Which brings us to crypto banking generally. These funds, protocols, businesses, whatever you want to call them and however conceived essentially offered the services of a bank - a return on unused funds combined with ready access to those funds when needed. Except, of course, when the markets started trending down the ‘ready access’ part disappeared fast, and we all discovered that the only innovation here was taking lots of risk with no plan for a downside scenario. Not great.

But before we get too depressive here let’s take a step back. The idea of a crypto bank is still a good one for the same reason the idea of a bank is a good one. Societally, we want to take money that is not being used and provide it to people that have uses for it. The person providing the money should get some benefits, interest, security, ease of access to their money, while the person using it should pay a reasonable price for the privilege. Obviously, the road between the saver and borrower is complicated, and maintaining that road is what banks do. It’s what crypto banks should do too. But, of course, the road ‘is’ complicated, and there is more to it than just moving money from here to there. So, if we dig a level deeper, what really are the problems that banks are solving?

Framing another way, how do banks make a profit? A hundred different ways, maybe more, but two of the most important and simplest to understand are the management of maturity risk and credit risk. What are these?

Maturity management is the classic way a bank makes money from time immemorial. The idea is that money lent for longer periods of time typically carries a higher interest rate - this makes sense, people are willing to pay more for the use of money for a long period of time. In a simple case, banks can take money from depositors (that is, you and me), which have essentially a zero maturity (you can show up any time to get your funds), and lends it out as a home loan with a 30-year term where the bank won’t receive its funds back for a very long time (they can’t call up the homeowner and say, we’d like our money back now, some of our depositors have shown up). If the bank can pay 2% on its deposit accounts, and charge 5% on its home loans it makes the 3% spread. Not bad.

But, there is a bit of a problem here. If the bank has lent out the money for 30 years, and you show up tomorrow expecting to take cash out of your account, that doesn’t really work.

This is not a new problem! It’s old as dirt! The solutions to it are sort of straightforward 1) don’t lend all the money out on long maturities, have layers at different maturities, 2) have at least some sources of funds that aren’t zero maturity (most banks borrow, often from each other or the public markets, at a variety of maturities), have layers at longer terms than just depositors who can show up any time, 3) have lenders of last resort you can access - ones that provide short-term funds to meet a temporarily high demand for withdrawals. Banks do all these things to try to manage maturities. What should this look like for a crypto bank? We don’t know because it’s a question we really haven’t asked. In light of all of the frozen accounts we are dealing with, we should be asking it now.

The other primary way banks make money is credit risk management. The idea is this - a bank needs to always make good on its own liabilities to depositors and other lenders, otherwise it is out of business. But on the other hand, many of its borrowers will fail to make their payments and there will be credit losses (this is not a maybe, it’s a definitely). If the bank has a big enough balance sheet, and is trusted enough, it can borrow at a lower rate than it can lend to more risky borrowers, and again collect the difference. The problem of course is getting the credit risk right so that losses don’t turn into insolvency. I hope you’re listening 3AC.

This is also not a new problem, and its solutions are again pretty straightforward. 1) don’t lend too much to any single counterparty or asset class, 2) make sure your rights in a credit default are as protected as much as possible, so you can recover as much and as quickly as possible if your borrower stops paying, 3) manage the pools of credit risk that you are taking and try to find the best risk/return areas in the marketplace.

And then of course there is the biggie, 4) maintain equity in the bank as a buffer for losses - these are the shareholders in the case of a normal bank - and are people that are willing to take the risks of banking and absorb near-term credit losses in return for being the owners of the bank benefiting if the bank turns a profit. A normal bank typically has a 10-15% equity buffer for this purpose, and a lot of people argue this is too low. What was Celsius’ equity buffer before it’s last raise? Approximately 5.3%. And we’re surprised to be where we are?

I hate to say it, but none of this is new. None of this is unusual, unexpected, surprising. Arguably, in crypto, where assets have high volatilities, and where a lot is still being figured out, crypto bankers should have been even more careful and conservative. Clearly they haven’t been, and we’ve all been taken along for the ride.

To be fair though, I don’t think we’ve yet seen any crypto ideas that allow us to address problems like these, which again, are the bread and butter of all banking. The approach so far has been - taking lots of risk because up only, while doing everything a normal bank does but worse, except this time crypto. But finding novel, crypto solutions to these problems is exactly what we need to do if we want crypto banking to be successful. To be the world changing thing we think it can be.

Until we do that we will remain in a mode of crypto finance that provides losses for many, gains for some, and yachts for the bank CEOs. And isn’t that basically just traditional banking?

Why are crypto bankers so bad at banking? For the last several weeks we’ve read as multiple crypto bankers, starting with 3AC, then Celsius, then BlockFi, and now potentially Voyager (which recently suspended withdrawals) have gotten themselves into trouble. It’s a bit of a problem if crypto finance is supposed to replace traditional finance as a way to provide credit and run our world’s economy. This isn’t exactly what we all had in mind.

The sad thing is that the errors being made are not new, novel or exciting (you’d hope that in crypto we’d at least be discovering new mistakes to make). They’re the same ones we have known about for decades (sometimes centuries) - bank runs, lax credit underwriting (along with its evil twin concentration risk), maturity mismatches, counterparty risk, contagion. Look at any of the recent bankruptcies, bailouts, freezes and the like, and you will find one or more of these concepts at work. And behind the concepts a CEO whose business plan was basically, up only.

There’s this idea of Chesterton’s Fence. It’s a worthwhile idea. The idea is this: you buy a large piece of rural property, many acres of rolling hills and lush valleys. One day, as you are walking around your newly acquired land you find a fence extending for miles in both directions across a beautiful valley. And you say to yourself, this fence is a bit of an eyesore, and really gets in the way of a nice walk. Should I get rid of it? And so this is the thought experiment - what are the arguments for and against removing the fence?

The arguments for removing it we’ve already made. It’s an eyesore, it gets in the way of a walk. What are the arguments against? Well, there aren’t really any, are there? You just bought the land, you don’t really know why the fence is there, you certainly don’t have a reason you can point to, do you? This is where Chesterton would step in and say there actually is a reason against tearing down the fence. The reason is that the prior owner was presumably a rational person, and when they put the fence up they probably had a reason for it. So, before removing it you should try to explore that past owner’s reasons, try to discover why they would want a fence there, and only after you have done that should you consider removing the fence.

I’m sad to say that in the crypto world we have often taken the approach of tearing down the fence immediately, and then being surprised by the often very good reasons that it was there to begin with.

Which brings us to crypto banking generally. These funds, protocols, businesses, whatever you want to call them and however conceived essentially offered the services of a bank - a return on unused funds combined with ready access to those funds when needed. Except, of course, when the markets started trending down the ‘ready access’ part disappeared fast, and we all discovered that the only innovation here was taking lots of risk with no plan for a downside scenario. Not great.

But before we get too depressive here let’s take a step back. The idea of a crypto bank is still a good one for the same reason the idea of a bank is a good one. Societally, we want to take money that is not being used and provide it to people that have uses for it. The person providing the money should get some benefits, interest, security, ease of access to their money, while the person using it should pay a reasonable price for the privilege. Obviously, the road between the saver and borrower is complicated, and maintaining that road is what banks do. It’s what crypto banks should do too. But, of course, the road ‘is’ complicated, and there is more to it than just moving money from here to there. So, if we dig a level deeper, what really are the problems that banks are solving?

Framing another way, how do banks make a profit? A hundred different ways, maybe more, but two of the most important and simplest to understand are the management of maturity risk and credit risk. What are these?

Maturity management is the classic way a bank makes money from time immemorial. The idea is that money lent for longer periods of time typically carries a higher interest rate - this makes sense, people are willing to pay more for the use of money for a long period of time. In a simple case, banks can take money from depositors (that is, you and me), which have essentially a zero maturity (you can show up any time to get your funds), and lends it out as a home loan with a 30-year term where the bank won’t receive its funds back for a very long time (they can’t call up the homeowner and say, we’d like our money back now, some of our depositors have shown up). If the bank can pay 2% on its deposit accounts, and charge 5% on its home loans it makes the 3% spread. Not bad.

But, there is a bit of a problem here. If the bank has lent out the money for 30 years, and you show up tomorrow expecting to take cash out of your account, that doesn’t really work.

This is not a new problem! It’s old as dirt! The solutions to it are sort of straightforward 1) don’t lend all the money out on long maturities, have layers at different maturities, 2) have at least some sources of funds that aren’t zero maturity (most banks borrow, often from each other or the public markets, at a variety of maturities), have layers at longer terms than just depositors who can show up any time, 3) have lenders of last resort you can access - ones that provide short-term funds to meet a temporarily high demand for withdrawals. Banks do all these things to try to manage maturities. What should this look like for a crypto bank? We don’t know because it’s a question we really haven’t asked. In light of all of the frozen accounts we are dealing with, we should be asking it now.

The other primary way banks make money is credit risk management. The idea is this - a bank needs to always make good on its own liabilities to depositors and other lenders, otherwise it is out of business. But on the other hand, many of its borrowers will fail to make their payments and there will be credit losses (this is not a maybe, it’s a definitely). If the bank has a big enough balance sheet, and is trusted enough, it can borrow at a lower rate than it can lend to more risky borrowers, and again collect the difference. The problem of course is getting the credit risk right so that losses don’t turn into insolvency. I hope you’re listening 3AC.

This is also not a new problem, and its solutions are again pretty straightforward. 1) don’t lend too much to any single counterparty or asset class, 2) make sure your rights in a credit default are as protected as much as possible, so you can recover as much and as quickly as possible if your borrower stops paying, 3) manage the pools of credit risk that you are taking and try to find the best risk/return areas in the marketplace.

And then of course there is the biggie, 4) maintain equity in the bank as a buffer for losses - these are the shareholders in the case of a normal bank - and are people that are willing to take the risks of banking and absorb near-term credit losses in return for being the owners of the bank benefiting if the bank turns a profit. A normal bank typically has a 10-15% equity buffer for this purpose, and a lot of people argue this is too low. What was Celsius’ equity buffer before it’s last raise? Approximately 5.3%. And we’re surprised to be where we are?

I hate to say it, but none of this is new. None of this is unusual, unexpected, surprising. Arguably, in crypto, where assets have high volatilities, and where a lot is still being figured out, crypto bankers should have been even more careful and conservative. Clearly they haven’t been, and we’ve all been taken along for the ride.

To be fair though, I don’t think we’ve yet seen any crypto ideas that allow us to address problems like these, which again, are the bread and butter of all banking. The approach so far has been - taking lots of risk because up only, while doing everything a normal bank does but worse, except this time crypto. But finding novel, crypto solutions to these problems is exactly what we need to do if we want crypto banking to be successful. To be the world changing thing we think it can be.

Until we do that we will remain in a mode of crypto finance that provides losses for many, gains for some, and yachts for the bank CEOs. And isn’t that basically just traditional banking?