 Who speaks about banking? All for that matter, and especially economists in economics should put up a warning like this. I'm going to be talking quite a bit about economics and economic theory, but mostly disparagingly, I'm afraid, because a lot of economic theory is, well, pretty bad. And I'm going to also advocate that one of the reasons for it being bad is actually that economics has been avoiding the banking system. And so we're going to talk quite a lot about the banking system, where it comes from. And hopefully, at the end of this talk, you will have a better understanding of it, why it causes problems, and at some level, how to at least anticipate them in your personal life. But what ultimately you're seeing here is what happens if a computer network tech engineer gets really interested in what is essentially a medieval network, the banking system, decides because it actually was a rainy weekend to sort of try to build a simple simulation of the economic textbook so I could understand it better. This is usually how I get myself into trouble and discovered that actually the economic textbook, the standard example, can't actually work as shown, which is kind of interesting to find in a textbook. So then I got a bit more interested in it and discovered how it really worked and talked to quite a lot of bankers who were surprisingly helpful and built what is, as far as I am aware, the first banking simulation and economic simulation that actually faithfully implements the double entry bookkeeping protocol that the banking system uses. So believe me or don't believe me, but I do think I have a fair claim to understanding how the system works just because I built a working copy. And you are welcome to go to the public version of it. We do have a development version that's quite a bit further forward. Get it working with the latest version of Java, download it, and play with it. And I warn you now, it's research code, so you may need to email me if you need some help doing any of that. But several long-suffering classes at Reconvict University have downloaded it, run it, and found out just how easy it is to crash the banking system. And it's at this point when I'm talking at economics conferences, I start getting very strange looks from economists. So we'll start talking a little bit about how the banking system actually works and how we reproduced it and why it keeps crashing. And just to sort of prequel it, I guess, what we did with the system I called Fred Needle once we got it up and working and had implemented double entry bookkeeping and could basically have multiple banks and look at the transactions between them was we isolated it. So the problem with any form of economic simulation is you need to create some kind of flow of money around the system. Banks create loans. They create money in the process. We'll get to that in a moment. But somebody then has to repay that loan, usually us. But companies take out loans all the time as well, so do governments. Governments quite often don't repay them. That can be a problem. And that money that you're using has to come from somebody. Well, it probably comes from a company. And that company has to hopefully be selling something for a profit to give you that money. And so you have to create these complicated loops of monetary flows. And that's quite difficult to do and still just figure out what is just the banking system's behavior and what is the economy's behavior. So we isolated the banking system very simply. By just creating a borrower who was paid a salary by the bank that matched the amount of money that they needed to pay their debt in any given simulation state. And we just looked at that. For a couple of years, all we did was just look at how banks behave if it's a completely closed system, not how it is in real life. And borrowers are repaying their debt, and we've played with different regulatory frameworks and so on, and saw what happened. So let's go back to where we began. The textbook description of the banking system, still in the textbook, still in Wikipedia, and honestly, Wikipedia gets worse every time I look at it, originates with the British economist John Maynard Keynes, who is writing a report to the British Parliament in 1931 on the big banking crash that Europe and the United States is just experiencing. And he gives this rather convoluted example. And he makes a couple of assumptions. And one of the assumptions he makes, which turns out to be very important, is that there is only one bank. So that actually gets rid of a problem, which we'll talk about in a moment, which is, you know, how do you transfer money between banks? That's actually quite tricky. That's, as soon as you have to transfer money between banks, you run straight into consensus problems. And anybody who's done distributed computing, maybe has heard of the Fisher Consensus Problem, it is impossible in that situation to guarantee that any two asynchronously connected nodes can agree on even a single bit value. And the single bit value in question here is your bank account. And people generally get very upset if that does not match what they want it to. So, you know, that problem has to be solved. So Keynes basically said, don't worry about that. We'll just look at this, you know, money creation thing that they do. And, you know, I'm trying to explain it to you politicians. Trust me, I'm an expert. So 60 or 70 years later, the textbook model of Fraction Reserve banking is sort of appearing like this. Sometime in the 60s, somebody in the Federal Reserve thought it would be better to take Keynes's single bank example and sort of split it out over a whole bunch of banks and basically hopefully to make it clearer. And so what he's showing with this example is the problem that everybody sort of screams about with the banking system, the money out of thin air problem. You know, banks create money, which they do. They also destroy it. This might actually be a larger problem. And banks create money every time they make a loan. So somebody took the Keynesian example, broke it out so that Bank A made a loan that ended up in Bank B. Bank B makes a loan where the deposited money lends up in Bank C and so on. And actually 20 years later and on Wikipedia, not in the textbooks, a computer scientist came along and said, oh yeah, that's a recursive equation. And it should stabilize to this. And so economic theory being misled essentially fundamentally by the Federal Reserve and John Maynard Keynes is predominantly based on the assumption that the money supply is stable. It might go through an initial expansion phase, but this is back sometime in the 1700s, 1800s. But now it's stable and so we can ignore it. We have a stable money supply. We know that supply and demand and market trading, which we've based all of our sort of systems on is a function of the money supply. I mean prices depend on the amount of money you can pay for something. Demand is expressed as money. And the supply, but we're just gonna assume the money supply is stable and that will be fine. And maybe if the money supply was stable, it would be. I don't know. We've actually run, we can run simulated economies with very stable money supplies. I'll talk about that in a moment. And we still have problems. So there's more going on with this than just the banking system for sure. But let's start with the thing that we can easily focus on. How do we know the banking system isn't stable? Well, Keynes didn't, or he kind of did, actually. If you look at the numbers in the back of the report, he's got about 10 years of economic data and you can sort of see that it's changing. But they didn't really have statistics in the way we have them now. Whereas now you can go to pretty much every central bank in the world and you will find either an M2 or an M3, a money supply series. And for every single country, it will look something like that, continuously expanding. But this is the fun part, expanding at different rates. This is a problem for the euro. Each banking system in each country, and sometimes in different parts of the same country, which also causes problems, it tends to expand its local money supply, its local lending supply at a different rate. Now, this is a very, very slow system. By human standards, it's glacial. All the interesting behavior in it is actually linked to lending. And loans can be quite long. A typical loan period if you're buying a house is 20, 25, 30 years. Go and look at the annual reports of all the companies these days. They've all got 30, 40 year loans on their books. So the behavior evolves over time with the lending and that takes decades. So a common theme in economics over the years and something that cryptocurrency latched onto is that we would like to have fixed currencies that have the same value with respect to each other because this solves a lot of problems. If I'm running a restaurant and I'm trying to import spices every month and I'm trying to price my meals accordingly, well, it'd be nice if those prices stayed the same. And of course they don't because foreign exchange happens and banking systems are expanding the money supply at different rates. So there's always these changes going on in foreign currency. So every so often basically economists try, they've tried with the euro and they used to be trying with something called the Bretton Woods Accord to fix currency relationships to each other across the different currencies. And this is okay for about 10, 15, 20 years depending on the expansion rates but ultimately it's destined to break and because they're all expanding at different rates. Think of a carpet. Think of a carpet where the threads of different parts of the carpet are expanding at a different rate. That's essentially what's happening when you're looking at different banking systems. And this is the UK. The UK was a three times a decade country. And I tend to measure it by how many times in a decade the money supply increases. The US up to recently was a two times a decade country. Stability is sort of like the two to three value. The Germans are very stable. So the Japanese, they're down at like 1.25. The Russians, well, they've had to revalue the rubble I think about six times the last time I checked. Everybody else is somewhere in the middle. It's quite interesting that from 2009 onwards the UK sort of stabilized and then is increasing again. I don't know what it means. I don't know if it was deliberate or not but it's changing something very fundamental in the entire economy. All right, so how does this actually happen? Well, these guys discovered statistical multiplexing. They discovered it essentially in the 14th century or so. Banking is based on a very old technology, I would say protocol basically called double entry bookkeeping. And in computer science words, double entry bookkeeping is essentially a single error correction detection protocol. So from the perspective of people who have to run companies and want to add up all the numbers and make sure that nobody's committed any fraud and that they've bought about the right number of things, it's a nice simple way to do that and correct any errors. And essentially the way it works is you take two people, you give each of them a book and each transaction the company makes, whether if it sells something to somebody, if it buys something, both people write an entry in their books. At the end of the day they add up the totals and hopefully they are the same. If they're not, at least the problem is localized. So the books are classified as assets or liabilities and equity and the operation is called a debit and a credit, except that the meaning of debit and the meaning of credit mathematically changes, depending on which side you're on. The way to remember this if you care is the right-hand side is what you think it is, credit adds and the left-hand side isn't. On the left-hand side, a debit adds and a credit subtracts and because that is done that way, it works. It's actually quite a nice system. So back in medieval times, the goldsmiths basically sort of started setting up and providing places to store gold for the rich people. Let's face it, it wasn't a large number of people who had gold in those days and they were using double entry book even to do this and so you would turn up, you would deposit your gold and you would get a single, basically a depositor account in there and they would give you a chit. It looked like this back in Sweden in the 1700s and that chit would basically say, I have a deposit of gold with this gold keeper. Now, here's the thing with gold. It's not like a pawn shop. You're not going in there saying, here's my shirt, I want my shirt back. You're just going in and saying, I want a certain quantity of gold back and I don't really care whose gold it was originally just as long as I get that back, right? It's perfectly reasonable. It's also the 1700s. It's 200 years before anyone is going to invent police services. So basically the rich people realized quite quickly I imagine that they are now going to a well-known location taking out a heavy valuable metal and trying to get away without anyone noticing. So very quickly, people stopped going to the goldsmith and start just signing these chits over to each other. So basically, you would give it to your friend, it's yours. And this very rapidly evolves into being cash money. Obviously, that makes sense. And meanwhile, the goldsmiths are kind of like, okay, well, we've got all this gold here and nobody seems to be coming in for it, right? Okay, I mean, let's be honest about this. They were probably the hackers of their time. What can we do? Well, we can make short-term loans, right? Why not? Nobody wants it. We'll just use our nice double-entry bookkeeping technology. We'll write 30 in the loan ledger and we'll charge them interest on that so we get something for ourselves. We'll make a liability deposit for them and we'll give them a chip for 30 more gold. And as long as not too many people come in on any given day and want their gold back, that's going to work just fine. And it does, actually. The problem, although it's not the popular view of the problem, the problem that, you know, a lot of people turn up and want their gold is actually not usually the real problem that's going on. I'll talk about it in a moment. But let me just draw your attention to something. As soon as you've got this set up and it emerges over about 30 or 40 years, you've basically got two forms of money, the sort of the physical kind that everybody can relate to. But in today's terms, on this side of the books, you've created virtual money. I mean, it's actually a huge argument with quite a few of the more obtuse economists. You know, is this money even money? Well, let's just say, let's just regard money as a token of exchange. Nothing else, store of value, forget it. As a token of exchange, absolutely this is money because what happens very shortly after the chits get invented is that the goldsmiths start taking checks. And basically, you can just write an instruction to the goldsmith and say transfer my money to my friend. And it all happens on that side of the book. So it is being used as money. There's no question about that. It doesn't leave the goldsmith. And so you've now got two forms of money in the economy. One is virtual and one is physical. And about 300 years of complete confusion for the economists. And it's statistically multiplexing. So it's gonna kind of work most of the time, but if too many people want their gold at the same time, it's a problem. Now, the one final problem they have to solve from a networking perspective for basically what is, if you just regard the money transmission side of this, a very elegant form of networking, I would say, is they need to be able to transfer deposits from one bank or goldsmith. But at this point in time, they are now becoming banks to another. And of course, when you do that, you run straight into consensus problems. Okay, so what the banks actually do is they transfer cash. So they take the physical form of the money and they deduct that from the deposit, they deduct the cash. And at least for a while, they physically send the cash over to the other bank. This is why if you sort of see 1960s, 1950s American criminal films, they're always robbing bank fans. That's what the bank fans are doing. Over in the British Empire, essentially you had a clearing meeting. The bank clerks would run around to all the other banks with the checks and things that they needed to balance up. Bank clerks are energetic young men. They weren't stupid either. They quickly realized that what they should do is all meet at a central location. So the Lombard Street Tavern, the second floor of which, they would all meet up there and they would basically all balance out the clearing between each other. And it took 50 years for anybody apparently in management to notice that all the clearing operations were being done down the pub. And when they did, they got their own premises. So you've got checks, you've got physical cash being used as a transfer, you've got mechanisms to balance up. You've got a network and you've got something that's actually very important for a modern economy, which is ways to transfer money between people that are relatively robust and allow commercial enterprise to occur. So what about all this lending? The thing that, you know, when you're watching the YouTube video, the vaguely anti-semitic tone usually starts to come in on. It's an emergent, you can see why they did this. I don't think there was any conspiracy involved. I mean, it actually makes sense. They sort of stumbled over a way to create a network. But the problem was, in some sense, it's embedded in the bookkeeping and you really need to understand the bookkeeping to appreciate what happens. When they create the money, they basically create a loan and they put the sort of money in the deposit account and when you repay the loan, this is what tends to get left out in a lot of cases, is, you know, what happens when you repay it? Hopefully you do. Well, the capital, the amount you borrowed, that is debited, sorry, credited because it's on the left-hand side and that shrinks the amount you owe and then your deposit account is also debited and that shrinks the amount of money that you have. It also shrinks the total amount of money in the whole system. So banks create money, sure, but they're also destroying it all the time and that's where the system can get a little bit quantum. And then when you pay interest, well, that just basically is a transfer from your account to either the bank's account or somebody else's account if it wasn't a bank that lent you money in the first place. So what goes wrong with this system? Well, what happens if you don't repay the loan? That's this last one. The bookkeeping is straightforward. You credit the loan, you write the loan off. Hopefully there was some collateral with that which you can take and sell and recover your money as a bank, but that might not work so well. You strictly then have to debit your profits. You basically write off that amount that you owed, that was owed to you, you're not gonna get it back unless you can sell the collateral. So you deduct that from your profits, okay? What happens if there aren't enough profits? And also, by the way, you've just destroyed some money because profits are also money, it's just the money that the bank earned and what you're doing. Well, this is fundamentally what causes a credit crisis. It's the economic consequence of an interruption in the credit supply from banks. I mean, it could be, there are other people in the economy lending but it's only the banks who interact directly with the money supply. And so if the banks stop lending and they stop creating money and the money supply starts shrinking, that has global and very unpleasant economic consequences. So the expanded accounting equation basically says that loan defaults are treated as an expense. So you deduct it from income and if you've got a really big loss then you're not gonna be paying any dividends and if that comes out negative then you're gonna start deducting from equity. Well, equity includes things like loss provisions and so on. So maybe that's not so bad. But fundamentally, as soon as this happens, as soon as you can't cover your losses with income, you start interacting with all the controls that over time people have put on this mechanism to stop it expanding the money supply too fast. Because we know what happens to a country where the money supply expands without any regulation, it's called the Weimar Republic, the Weimar hyperinflation, the Zimbabwean hyperinflation, the Icelandic hyperinflation of 1970 to 1982. It basically, that's actually the quickest and fastest way to destroy the entire financial system that anyone comes up with and the social consequences usually afterwards are disastrous. Okay, so how bad can it be? How much loan defaults can the banking system absorb in a year? Well, and that's the problem. And the math is surprisingly easy to do. So I don't know why this isn't pointed out more often, but it doesn't seem to be. Banks profits are essentially, take the total amount of interest they can get on their loans, take the total amount of interest they have to pay on their savings accounts. So banks don't really make profits in the way that a lot of companies do. They live on the spread between lending and savings. So this can vary a little bit, but 4% is not an unreasonable figure for how much on average a bank is making in terms of interest. Generally banks, they might charge one person 10% because they think it's a really risky loan. They might charge a company that's really stable only 3%. It also depends on what the central banks are doing the interest rate, but they care about the spread and the spread will be about 4%. So back at the envelope calculations, assume that 3% was gonna cover salaries, dividends, big bonuses at the end of the year, nice parties and so on. Well, you've got 1% to cover your losses. The industry figure is 0.75%. So we have a critical system, system that our entire economy depends on, a system that if it breaks in any particular way will destroy an entire society and has done on several occasions. And its fault tolerance is about 0.75% every year. Great. Would not pass design class. I wouldn't, but it never got designed. It got emerged. Humans being humans, doing what we do best, try it and see. So you can, depending on the country, look at sort of some of the evidence behind this. Now I'll talk about this again at the end of the talk, but a continuing problem which you see with a lot of economic figures is they don't factor in any of this money supply expansion stuff, right? So this is the chart for how much net write-offs US banks have done over the last 30 years. And when you're looking at it, you have to kind of squint and go, okay, 30 years ago the total money supply was about a sixth what it is now, right? That's the money supply. Back in sort of 1986, it's just past the two trillion value. And so as a proportion, a $2 billion write-off level is actually a much higher percentage of the total money supply back in the 1980s than it will be the following decade and especially now. But you can see that these, basically the driving mechanism of recession is these losses being very low and then suddenly increasing above the point that banks can absorb them. Now it depends a little bit on how bad the situation is, what your central bank will do. In the 1980s, what I've saw somebody say at an economics conference one time was basically everybody knew, everybody being the central banks around the world, that the entire US system was insolvent. I mean, it's a bit of an exaggeration, but it was kind of what was going on. They'd made a lot of loans to South America and the South American governments had all defaulted on it. And actually governments defaulting on back loans has historically been one of the big issues with this system. Because you can't do anything about that when you're a bank here. You can beat up the little guy, but there's not a lot you can do with Charles II. So basically everybody sort of ignored that, made whatever tolerances they could and eventually the system got to the point where it could drain out the losses from the system and it was pretty stable for the 1990s and then the internet bubble happens and something else happens which is called loan securitization. I'll talk about that in a moment and that causes a huge wave of defaults in sort of the 2000s now. Loan defaults are what we would call a lagging indicator. Well, that's what the economists would call it. I would call it latency problem. Somebody defaults on the loan, they stop paying their interest, they probably stop paying the capital as well. The bank doesn't just then and they're right it off. In fact, if the bank can't afford to ride it off without hitting its bank capital, it's gonna keep that loan on its books. This is the so-called zombie bank phenomena where a bank has a lot of bad debt, it's making a little bit of money on the interest perhaps and it can't make any new loans because the rules won't let it. Common in Japan in the 1990s and also in Italy now. So the banks will sort of bleed off these excess periods and that's why you see this sort of net write-off thing actually lagging the main event of the recession as well. So in 2008, the credit markets froze in 2008, but the write-offs didn't really peak until a couple of years later. So one of the problems in the entire system is latency and various accompaniments for that. So to understand what the central banks are gonna do in any given situation, basically they will act to protect the banking system and rule zero is money must not shrink. Rule one is that debts must be repaid because if they are not repaid, the money supply will shrink and rule two, which is what the politicians have told them is inflation should be less than 2% a year. Only they don't really have any way to do that but we'll ignore that problem for a moment. Why must the money supply not shrink? Well, because the last time it shrunk in a big way, it destroyed the entire world, not exaggerating there. You can draw a straight line between the US banking system essentially collapsing and the money supply in the US shrinks by a third in the Great Depression. Now you'll still see economists saying no, no, no, it didn't. Really, it was the Fed's fault. Actually no, it wasn't. When economists say the money supply didn't shrink in the Great Depression, what they're talking about is that asset money, the semaphore money, the cash money, that actually does stay the same. It's the liability money that shrinks as bank after bank in the Great Depression went bankrupt and was just wiped out, the money supply progressively shrunk. Well, in a market-based economy, prices can adjust in that situation. Prices will go down, salaries will go down. That is painful, but it is an adjustment. What doesn't adjust is the critical thing. What doesn't adjust is debt. We have no good way to adjust debt repayment especially from the banking perspective because the banks are stuck there. If they change the value of the loan outstanding, they've also shrunk the money supply. So they're stuck with this. And so as the 1930s progressed, more and more banks failed, more and more companies were unable to repay their debts, more and more banks failed because of that and the entire system basically reaches bottom in 1933 when Roosevelt intervenes, revalues the value of gold and kind of kicks the banking system off again. It was probably about to be do it by itself anyway. I mean, the system is somewhat resilient but the social consequences are enormous. One of the social consequences was that during the 1920s, the New York banks had made a lot of loans to Germany to help the Germans recover, the Weimar government then. And in 1929, they called them all back because you can do that sometimes in the system. You can just say, we pay me now, all right? Of course they couldn't, so the German banks failed. Then there's a cascade problem in Germany. Hitler's performance, the voting for Hitler was actually going down a little bit in 28 but by 1932, 1933, with the German economy collapsing again, second time in 10 years essentially, right wing progressivism seems like a good choice. And the rest, as they say, is history. Okay, so it's not a very stable system. It's got terrible fault tolerance. Just for the record, is it possible to have a stable banking system? Well, yes, we can run a simulated stable banking system under very peculiar conditions which would never exist in the real world. All account transfers have to occur within the same bank. That's an interesting one. If there are unbalanced flows between the banks, that creates what we call liquidity problems. One bank doesn't have enough asset cash to pay their depositors and another bank has too much. And actually loans will do this all by themselves. I'll show you in a moment. Loan defaults have to stay below the magic profits and expenses thing or that might be possible if bank was very, very conservative about its lending. And this tends to be how banks do behave. We'll run that once anyway. And it's also why banks have progressively become far keener on lending on property which has a value as a collateral that tends to retain itself than for the original reason they were set up which was lending to people to create businesses which is also a problem. And if you want to control the money supply, if you want to have a completely stable money supply that is actually possible. You just have to have a strict asset cash reserve requirement and you have to apply it to all the deposit liability accounts and not just some of them. And if you look at the American banking system, for example, it has a 10% reserve requirement but only on certain kinds of deposit account. So, you know, there's that. But it is possible and it can be done. So that raises another interesting question. Should the money supply actually be constant? It's a good question. Economists and I actually kind of sort of agree with them. Think not. They think that there's some value in having a low rate of inflation. Not a high one, that's problematic. We're about to find out how problematic but a low one kind of eases the invisible constraints on people repaying their loans, collateral being perhaps a little worth a little bit more every year. So if there's a loan default to the bank, it's not quite as serious. And it also means, funnily enough, that there's actually more new lending. So if you look at the amount of new lending, you can have new loans, assuming you've regulated the system to not expand too fast. Can either come from somebody repaying a loan so the bank's got a little bit more lending that it can do to somebody else, or it can come from the expansion of the monetary supply. And if you look at the numbers, actually, for especially short loans and a reasonable monetary expansion rate, the US is about 7% a year, or at least it was. More new lending is gonna come from monetary expansion than it is from people repaying their loans. And if you need new lending, well, that's something to consider. Why might you need new lending? Well, as I said, there are a lot of problems with banking, some of them are economic, some of them are systemic. One of the economic problems is this, very small rates of default can break the whole system. Who's a big borrower from amongst other people, the banking system? Well, in the US, power companies. Power companies are building big power plants and these have to be financed over 30 or 40 years, so there's a lot of lending to the power industry. The coal power industry, and some of the nuclear power industry as well. Now, it would be very sensible to switch all that lending and more over to natural forms of power generation, silicon especially, but also wave and so on, and to do that now. And I think people would, except that it would crash the banking system. It really would. There's more than enough debt in those power companies to take out the banking system in a big way. And the interesting part is that might just happen anyway. I mean, the improvements in solar cells are accelerating at such a rate that this might just happen without anybody trying to get involved or try to stop it. I hope they don't try to stop it, but you know. So, there is this element too to banking that sometimes the banking system goes wrong, the regulations fail or something bad happens, and sometimes the economy goes wrong. Something happens in the economy like, you know, solar cells, or in the 1920s, it was basically the invention of the Combine Harvester. Speed up that part of the economy. It doesn't necessarily need lending anymore, and the whole system gets out of whack again. So, you know, we don't know really enough about that side of things, but it's definitely a factor there in history. And we see this a bit in the Great Recession in 2008. What caused the previous crash? Because we're now into another one. So, another element of the history of the system, the long one, I could give a whole lecture on this, is how do we control it? And first people try to control the expansion rate to stop, you know, the money supply expanding too much. This was done naturally by depending on the price of gold, which it works up to a point, but actually every time somebody writes a check and uses that virtual money on the right-hand side, they influence the price of gold without the sort of gold itself being involved. And so, if you look at the figures for the gold standards, there's sort of been two periods when we used gold that way, you actually see a slow, steady expansion of the money supply, even in this period where a lot of people would have you believe in this table. It wasn't. But it wasn't expanding that fast, so it kind of worked for a while. After the collapse of the second gold standard, which was the Bretton Woods fixed currency exchange scheme, and fundamentally it collapses because everybody's buying oil from Arabia, huge amounts of money are piling up there. There isn't a good way for it to come back. All the banking systems are sort of expanding at different rates, and eventually it became possible to arbitrage the price of gold between London and the US, and that lasted for about three days before Nixon was told to pull the US out of the Bretton Woods scheme. For the next 10 years, the system kind of freewheels, which causes a lot of countries immense problems, and Iceland in particular has a hyperinflation because of that, and then the Basel capital controls emerge in the 1980s, and Basel capital effectively shifts regulation from the reserve requirement on the left to the capital on the right, and capital is a few things. Unfortunately, one of the things capital can be is some forms of debt, but mostly it's things like lost provisions and the initial injection of cash that allows the bank to be a bank in the first place. So, okay, that's fine. It actually works quite well as a regulatory framework. Problem was, way back in the 1930s, and as I said, it's a very slow system, so sometimes the causes of problems are really removed from the sort of final denouement, as it were. The US has a great depression measure brought in something called loan securitisation. Loan securitisation basically allows a bank to take a loan that it's made, create some money with, and just sell it to somebody else. I mean, it's a decent investment loan, a treasurer's loan. So, Fannie and Freddie May did this for decades at a very slow level. It didn't do anything particularly bad. It probably helped the economy. It slightly increased the total amount of debt available, and then computerisation comes in in the 1980s and 1990s, and it suddenly becomes possible to do it on large scale, and Salmon Brothers is actually the sort of company that kicks this off, and suddenly there is a lot more debt in the system, because up till that point, with traditional fractional reserve banking, without it really being explicit, there was always this one-to-one relationship between the amount of debt and the amount of money, and loan securitisation broke that completely, and suddenly you had a lot more lending and a certain amount of money. Think of it as a water system. You've got pipes, the pipes are loans. Money is flowing around them. You increase the total number of pipes, but you leave the amount of water alone, that's what loan securitisation is, and suddenly the pressure starts going down, and at some point in the system, there isn't any water flowing at all, and that's where the loan defaults start, and basically that was the 2008 crash. It takes a while to build up, because fundamentally debt is an asymmetric network flow. By asymmetric, we mean that more is going in one way than another, and this is where I came in as a network engineer and go, okay, that's interesting, because in computer networks, when we have that going on, it usually causes quite a lot of problems, and it does in banking networks as well. The loan securitisation broke some sort of things that the banks were doing without even really telling anyone. One of the things banks do is try to keep all these money flows on their books. They would rather, there wasn't any interbank flows. They want their lenders to basically be basically perhaps working for a company that's also borrowing from them, and so all of this money just stays on the right-hand side, no cash involved, everybody's happy, all right? And they will make decisions according to this, and quite often you'll be told, if you want to borrow from a bank, you need to have your account, you need to have your salary paid into us, and that's why. Okay, loan securitisation just let them sell these loans willy-nilly across the system, so set up all these cross-bank flows, and after a while, what happens basically if you've got your borrower here and they take their loan over to some other bank? Well, they're paying back more than you originally gave them, and bank three starts losing its cash. Without bank three of any control over this, bank three is seeing a continuous flow of money back to bank two, basically the interest rate. This is one of the ways that interest rates affect the banking system, which is not acknowledged. The higher the banking, the higher the interest rate, the faster that process will naturally be, so it could be about to get interesting. And what you also saw in this period, and this may not be that visible, I apologise, banks started taking loans from other people, so they would also borrow and re-lend, and that doesn't make the banking system any stabler. So, cut a long story short, the United States printed a whole bunch of money, used it to buy the bad debt on the banking books, put it aside, kept it away, and actually did manage to more or less recover the system. So, and this is, we're showing here is M zero. And M zero is the asset, the physical cash that we keep talking about. M two is the deposit account, and you can see this jumping up and down like a yo-yo, but the M two, up until a relative recently, was pretty stable, expanding but stable. So, what happens in 2020? Why do we currently have inflation? What aren't the central banks really wanting to admit in public, although if you read the right articles in the financial times, they do admit it. In 2020, with the entire world basically going, let's shut down the economy, which is not something that's ever been designed to do. Essentially, it was obvious to everybody that within a month or two, the entire banking system would collapse because nobody had any salaries to pay their debt. So, without anyone even almost discussing it, the American Bank basically printed $2 trillion and handed out to everybody. And unlike the previous time, when you give money to everybody in the system the way they did, that's a direct injection into the banking system, and the banking system then starts to expand that quite quickly, even under credit capital reserves. And you can see here that, whereas it was a sort of two times a decade system, now it's expanding a bit faster. And the second expansion isn't really the government so much. This is the banking system getting going and making more lending and more money. So, that is today's problem. Now, whether it's stabilizing or whether it's going to continue expanding, I don't know, the EU looks the same. Most countries are starting to look the same. We are probably looking at a period of increased inflation and increasing interest rates probably isn't going to help that much with that. To actually stop this, you need to intervene with the lending itself and sort of tell the banks, okay, you need to basically be, I would just put rules on. I would say, okay, you can lend to new forms of energy. You can lend to somebody who's never bought a house before. That person over there who wants to refinance their home and buy a Mercedes or something, well, tough luck. But that's just me. All right, that was the crash course on banking. I've probably confused everyone. How can we get involved with this? Because I can tell you now, economists, honestly, I'll skip through all this, sorry, if you want to see it, it's interesting but we're out of time. Economic models and they have them are terrible. I mean, it's beyond pre-copernican. We are talking, the best way to describe existing economic models is they are trying to fit a very short ruler to a very long curve. And it's even worse than that because there's a whole set of papers basically called calibration, which is how to get your model to fit the data that you're getting. And yeah, if you take a short ruler and you've got a system that's sort of doing this, you can fit it. There's actually more doing this to be honest, but anyway, you can get it to fit the data for a certain amount of time and then, you know, it stops. They're not gonna get out of this on their own. They really aren't. They're trapped in basically textbooks that are presenting them with falsehoods about how the whole system works, a very complicated system and no tools to analyze it properly. So I feel quite strongly that we sort of as computer scientists as networks, especially as programmers, and it is quite fun actually to build economic simulations and crash the banking system, need to basically provide them with tools and there's a huge amount of work to be done. And if you want to be sort of, you know, talk to me about how perhaps we start organizing this, please catch me after the talk. By all means, play with Fred Needle. If you want, if it doesn't work for you, call me, I'll get you up and started. It is free for non-commercial use and if you want to do something commercial, we'll figure out some way. Other than that, one thing we could all do, and I would advise you to do it just personally, figure out for the country you're living in what your monetary expansion rate is. Is it two times? Is it three times? Is it four times? Whatever. And then start a movement to correct all financial numbers for expansion in M3 and especially government debt because there's this constant thing of, oh my God, the right way when people say, the government is spending more and more. And it is because there's more and more money in the system every year and so they get more and more tax income because of that. They're fairly hooked into the whole cycle. And yes, they're spending more but prices are going up and if you actually look, for example, at how much the US government is receiving as a percentage of the total money supply, it's getting smaller every year. And I could give a separate talk on why that is and it's essentially because money is piling up in the financial system and can't escape into the other system. So we get more and more rich people or actually we get fewer and fewer but they're more rich and the government's getting actually less money to spend every year as proportion of the total money supply and nobody's really sort of thinking things through in that level. There is a fair amount of correction of, they will correct for inflation. Well, that's fine but the thing with inflation is we're producing more and more every year as well. So it's a bit of a misleading statistic. It's a function of both the amount of stuff and the money supply. So I think just correcting by the money supply is the way to go. All they correct by GDP and yes, that's another slew of problems. So at any rate, if you would like to be involved in a sort of open source movement to try to develop economic models that are actually based on real economies and not fantasies, get me after the talk. Thank you very much for listening. If anybody has questions, please move to the mic. One question per person, a few minutes left. So what are generally the differences between Threadneedle and this Minsky software that Steve Keane talks about a lot? Yeah, the basic difference is I use double entry bookkeeping. There are several models, Minsky is one of them, that are based on sort of the balance sheet view. And that's the level above. And that would be great and I think it would work well. It does actually, it's better than just anything else. The problem is that there are some things that can, quite important things, that can be different at the double entry bookkeeping level but the same at the accounting level. We ran into this in Iceland, funnily enough. Negatively immortalized loans, which is a form of loan where the capital grows rather than shrinks. The double entry bookkeeping there really matters in terms of how the system behaves and yet the accounting view simply doesn't catch it. Double entry bookkeeping is uncontrovertible. It's a protocol. It's just like TCPIP. We can audit it. We can be sure of it. It's the atomic level of banking system. I would strongly recommend we go with that. But that's between me and Keen, I guess. Hey, I get the impression that most of the problems you discussed here are entirely a function of the fractional reserve system where banks loan money they don't actually have and that if you would require them to only loan money they actually have, most of it would go away. Is that true? And if so, why is the system necessary? Well, it's a very good question about the necessity. You would get a different set of problems, is the evidence we have from simulation. Why is the system necessary? Well, it's really, really hard, and the crypto people have been discovering this, to scale to the planet, right? To have a monetary transfer system that allows you to go from Iceland and go into an ATM in Tokyo and take out money is no small achievement. And the system you've just described probably wouldn't be able to do that. And it's hard to prove these sort of things, but I will point out that crypto essentially reinvented fractional reserve banking, right? They did. You look at Teva, for example, it's got loans on the asset side, and they did this very, very quickly. So it's actually an interesting question is, can you stop people reinventing fractional reserve banking, right? Because it's statistical multiplexing, and actually most networks, when you get into them, your ISP, it basically takes a 10 gigabit per second line and it sells it to 100 people, one gigabit per second each. That is a form of fractional reserve banking, or statistical multiplexing. It works until everybody wants to stream the same thing at the same time. So if you can design a network system that scales, that meets fishy consensus problems, and that separates completely money and banking, I think everybody would be a lot happier. But until we can do that, we're stuck with it. Cryptocurrencies exist, and assume that Bitcoin can scale to be money, which you currently can't, but assume that it gets there. What effect does that have on the existing banking system? I think that would probably replace the banking system. But let me point out, and I've published a paper on this, although I couldn't get into a journal because they wanted me to solve the problem, it is provably impossible for Bitcoin to scale to that kind of level. It's based on a full mesh network, and the scaling properties of a full mesh network are order n brackets n minus 1, number of nodes. You can't scale Bitcoin to anything close to what will be required. And that's why people have reinvented fractional reserve banking with it. At some level, it's a really elegant solution. And then you look at all the problems that come with it, and you go, oh my god. On a scale of 1 to 10, how badly are solar panel leasing companies going to screw us over? I have no idea. Honestly, I'm sorry, I don't know anything about that. But it's going to screw us over a lot more if we don't do something about global warming. Right?