 If you were just to look at market value of equity of banks and the book value of equity of banks, these are two statistics that are entering into economic models. People are not really agreeing on which one should enter those, but they diverged completely during the crisis. It's just an aggregate statistic, but it tells us a lot about what's going on. I mean, if you just look at the book equity, it tells you that accounting values often mask risk exposures and even realize the risk and the crisis. And maybe also market equity sometimes overreacts, but at the same time, they were in fact much earlier on catching up to what's going on in the banking sector than book equity was. I'm Juliana Begenau. I'm a system professor of finance at Stanford's Graduate School of Business. So my research is generally really focused on banks and banking and the measurement of risk and banking. There's a lot of methods in trying to figure out what's the risk of banking and I believe the approach that I bring here is both novel and also applicable to particularly the question that our banks are facing and the type of risk that they are taking. What is often a concern is that regulators and academics often are lagging behind, whereas the new risk that they are taking is spotting the risk. You see that a lot of times that the core reports, these are kind of the regulatory findings, often introduce new variables after the fact, which is very difficult for us because that's kind of the data that we are using as academics in order to figure out what type of risk that they're taking. So one needs to have a fairly in some sense precise but a broad approach in order to figure out what are actually new risk and where they could evolve and where to zoom in on banks actually and what are the risks they're having. So it's a kind of a well-tested method. It's called kind of the replicating portfolio approach. A lot of financial management firms do that. That's what we teach also our MBA students, kind of how to value companies while you're kind of figuring out what are the cash flows that these companies have and what is the underlying risk and then you're trying to figure out what's my best alternative to these cash flows and that allows us to get a sense of what is the risk of those cash flows. And that's in the same way as that's how I approach banks. Banks have in a sense, I mean they have a very complicated portfolio. What we are trying to do and I'm trying to do is kind of both to reduce the dimensionality of this complicated portfolio and to map it into something that is again that we can make sense of and understand risk exposure. For example, if you're thinking of the banks, although complicated portfolio in one hand but they do have a lot of securities which are bonds, treasury bonds and loans and those can be mapped fairly easily into a credit market portfolio that has both the same term exposure, the same interest rate risk exposure that banks have and the same credit risk exposure. And you can verify that the cash flows of those both of those portfolios look very similar and from there you can then deduce well the same risk exposure that this capital market portfolio, this replicating portfolio have, it's the same risk exposure that banks have. By comparing basically the cash flows of the banks and what the regulatory reports tell us what is like the position, the amount of the position, how much interest rate exposure is there, we can kind of deduce what would be the best alternative capital market object that is a kind of a bond portfolio and from there we can deduce the risk. So other part of the equation is leverage, right? So as anybody that works on credit is knows banks have a lot of leverage much more than other financial institutions. So even if the asset side doesn't look super risky, if you lever it up a lot, even though the loss is like 1% but you lever 10 times and it's like a 10% loss that could hit you. So that's always something to be worried about in the case of banks. I mean if interest rates rise and they do have some notable interest rate exposure, this will make the value of the portfolio decline. Regulators trying to keep tabs on a lot of sorts of risk, it's hard to predict what's going to be. So I think we are as an economist we are more tend to be a doctor to diagnose what the illness was about and maybe to think about treatments and make some preventative measures but it's really hard to anticipate who else is going to get sick and what will be the next illness that is going to afflict the banking sector. I think leverage is always a concern. I think we are more general concern is to assess really the risk of banks because a lot of their data comes in terms of accounting values. For example, net interest margin is one of the key performance statistics of banking. It means like interest income on assets minus interest expense on debt. This net interest margin has been very flat over time and then it has been concluded that there is not a lot of interest rate or even credit risk exposure. Cayman Times' interest rate risk was high in the 80s particularly when we had these sudden interest rate spikes in the end of the 70s and we had also credit exposure crisis happened in the 2008-2009 crisis. It's concerning that risk management of banks is not that they don't use an economic measure but in some sense non-fundamental partial revenue measures in order to figure out what their risk exposure is. It's not an entire story but that's their answer to do we have interest rate risk? No, our net interest margins are flat. I've also studied the cross-section of banks before the financial crisis. It was clear that those banks that had a lot of leverage and fairly high asset risk by the crude standards that regulators allow us to categorize banks into it was possible to fairly accurately predict which are the banks that are going to lose more in the crisis than others. So there's a lot of studies that have shown also that the same banks that have unrisky behavior in the earlier crisis continue to do so. There's both, I mean there's a predictive nature and there's also kind of persistency and like the behaviors that lead to further crisis. We spent a lot of time in justifying exposed why banks are so special, why they cannot be evaluated with our standard battery of tools and if banks were special we shouldn't be able to replicate their cash flows with our capital market portfolios which kind of mimic the risk and the cash flows of a banking sector. So we should do very bad. One thing that banks are special in, they do have monopoly power over the payment system and that is a unique role of banks but it also comes at a cost, I mean operating cost of running this payment system is fairly high. So even if you were to fund yourself at the deposits which gives banks a rate advantage but you also have to provide as a bank the brick-a-motor, the ATM network so far, I mean as long as it's not totally in the online world you'll need to do that. Recently there's been a lot of push towards understanding micro movement that can be causally interpreted although I think that's very important and should definitely be done. This has come at the expense a little bit at this view of like all aggregate studies are somehow wrong or not really helpful because we can't really make any causal assessments and we do can actually learn a lot from aggregate movements and aggregate studies and from just like looking at aggregate data. So if you were just to trying to explain what banks risk exposure is based and what their portfolio looks like based on regulatory information, we can do this pretty well, we can describe position by position of banks quite well then if we bring it all together aggregate and look at what would be our implied, the implied equity position of this replicating portfolio and compare that to the actual bank stock market data, diverge completely. So it doesn't look a lot alike and then you can say oh well maybe we did a mistake but then why does position by position work well or does the stock market think about banks in a different way. That's another also methodological issue so if you use stock market data and trying to understand what banks do and where the risk exposure is which is also a common technique and running stock return regressions and factors and figuring out where does the factor sensitivity, you will find that banks have no credit risk and no interest rate risk exposure. So what is going on? Why does do stock market participants not incorporate that thinking? And also another interesting fact, a lot of stock market shareholders in general seem to have valued banks more the more levered they were before the crisis. Although then they exposed, I mean suffered the highest losses so I think this will be an important thinking about systematic ways of behavioral or inefficiencies that can come from a source from some behavioral mechanism will be hugely important I think for economics.