we set up groups to oversee those AAA securities used in repo transactions, and perhaps put a government guarantee on them

TO BE NOTED: From Falkenblog:

"Gary Gorton Explains the Crisis

This financial crisis has had two legs: the initial boom and bust in housing prices, primarily in the USA, and the accelerator mechanism that then turned this crisis into a panic, affecting financial institutions almost indiscriminately, and countries furthest away from the USA (eg, Russia, Brazil), actually were among the worst performers. I think the boom and bust in the housing market was a combination of a benign history combined with a variety of forces encouraging lower mortgage lending standards under the assumption that a disparate origination statistics by race were the result of arbitrary underwriting standards, and that housing prices would not fall in aggregate (nominally, they had not really done so in the historical record). The government keeps meticulous data on loan originations by race to satisfy the Home Mortgage Disclosure Act, but does not keep track of default rates by these same categories (see here). After all, you don't want to give lenders an ability to defend their unconscionable redlining.

But until today, I was totally befuddled by the transmission mechanism, the amplifier. I mentioned that last year at this time I was at an NBER conference, and Markus Brunnemeier pointed out that the decline in housing prices was insignificant relative to the change in stock market, and this was in May 2008! Almost everyone in the audience of esteemed financial researchers agreed, and thought the market was in a curious overreaction. That is, we did not know why a loss of $X in housing valuation, caused a $10X change in the stock market--and it was only to get much worse. You see, greed, hubris, fat tails, copulas, regulatory arbitrage, do not really work. Merton's explanation that as equity prices fall, both equity values and volatility increase, presumes the equity value fall that needs explaining (after all, it is much greater than the value of the housing price fall).

Gary Gorton makes the first really compelling explanation I have seen, and he's an economist. He is drawing on his experience as a consultant at AIG, and his experience modeling bank runs. I think this highlights that those best able to fix things should not have totally clean hands. Banks are complex institutions that have a lot of very particular attributes to their balance sheets and cash flows, and while a physicist or non-economist would be untainted by failure to foresee the crisis, this also neglects the steep learning curve they would also face. It is easier to correct experts than create them anew, groups are rarely incorrigible when it comes to big mistakes everyone sees as such (in contrast, to say a post-modern literary theorist, whose failures are less concrete). Gorton did not anticipate the 2008 events, but his analysis is much richer than those blaming complexity or hubris (Merton did a nice job of noting that ABS are actually much less complex than your average equity security given that it is a residual to a firm with lots of discretion).

The Gorton story is as follows. We have not had a true bank run since the Great Depression in the US, so we forgot what they look like. The essence of a bank run is described in those books on Bear Stearns: sparked by plausible speculation and a first mover, if you think a bank is losing other depositors, it pays to take out your deposits, and as no bank is sufficiently liquid to pay off all its depositors immediately, if this idea become popular the bank is doomed as each depositor seeks its own self-interest in getting to be first in line for their money.

In the old system, retail customers would put their money in a bank. These depositors have assets that can be used for transactional purposes, they can write a check, and have the money wired out in a matter of days. Yet, the money can be lent by bank, and using the law of large numbers, the bank can estimate their daily cash needs, keep a sufficient amount of reserves, and everything is kosher. Since the Great Depression, retail runs are no longer an issue. Yet now financial institutions are highly dependent on wholesale depositors. These are institutions, like hedge funds, and instead of depositing money in a bank, they deposit Treasuries, and more frequently, AAA rated securities (often derivatives). These depositor have overnight repo returns that can be used for transactional purposes because they can be turned to cash in a day's notice, but just like the retail story with the money, the AAA bonds the wholesale depositors leave can be lent (rehypothecated) by the bank. Retail cash was replaced by AAA debt.

Thus AAA rated bonds become cash used as deposits, and as AAA rated Asset Backed Securities had higher yields, but similar treatment to Treasuries, they gained more and more popularity. A key to these AAA rated bonds is that their default rates are, basically, zero. That is, at Moody's we estimated them at 0.01% annually. One should expect to see one or two of these securities default in one's working life. Thus, they are informationally insensitive, because they have so much safety in them, going over their risks is pointless, uninteresting. That is, as these things have reams of data that show zero default rates, and the market thinks they have almost zero default rate, it was counterproductive to get worked up about their future default rates. You would have wasted a career crying wolf about AAA-rated securities from 1940-2005.

A systemic shock to the financial system is an event thtat causes such debt to become informationally sensitive, in this case, the shock to housing prices. But when these assets become stressed, everyone is understandably perplexed. It's as if the electricity stops working, an event no one really contemplated, and it is too late for everyone to become an electrician. Thus, the concern about asset backed securities and AAA rated bonds of all kinds, become suspect. What were previously information insensitive becomes really complicated, as one works through the details of derivatives, or balance sheets of banks, that were previously assumed to have a 0.01% annualized default rate. This is because bayesian updating suggests that when you have several AAA rated defaults, the probability is that your model is incorrect, as opposed to just having bad luck, and you have to first figure out what that model is, and where it broke down (ie, is it just the housing price assumption?).

This is consistent with bank panics pre-1933 in the USA. Historically, the public becomes wary of all bank deposits, and the bank's balance sheets go from informationally insensitive to sensitive, unmasking their complexity. Nonetheless, the actual number of bank failures in these crises was comparatively low (excluding the Great Depression).

Say the banking system has assets of $100 financed by equity capital of $10, long-term debt of $40, and short term financing in the form of repo of $50. In the panic, repo haircuts rise to 20 percent as banks are wary of previously Gold AAA rated securities, amounting to a withdrawl of $10, so the system has to either shrink, borrow, or get an equity injection to make up for this. As we saw, after some early equity injections during the fall of 2007, this source dried up, as did the possibility of borrowing. That leaves asset sales. So the system as a whole needs to sell $10 of assets. They sell their tradeable securities as their liabilities decrease, and those securities are in a pricing matrix with the AAA rated collateral that is increasing its haircut. Gorton estimates the banking system needed to replace about $2 trillion of financing when the repo market haircuts rose. This is the amplifier.


The increase in haircuts basically removed liabilities from the banking system, and the prices fell on many comparable securities as banks shed assets. In the fire sale, assets are now worth $80; equity is wiped out. The system is insolvent at current market price levels.

How do we know depositors were confused about which banks were at risk? The evidence is that non-subprime related asset classes like auto loans, credit cards, and emerging market debt saw their spreads rise significantly only when the interbank market started to breakdown.

A banking panic is a systemic event because the banking system as a whole is always insolvent in such a scenario.

Thus, the problem is that repo replaced retail deposits as the center of the bank run, as AAA rated securities of all types used in these transactions were all downgraded implicitly to non-prime (below A2/P2), and in many cases based on Mortgages which historically were the safest asset class in the US. There was a bank run, but it was by the 'shadow bank sector', and so all we could see were the effects, but not the cause.

His suggestion is that we set up groups to oversee those AAA securities used in repo transactions, and perhaps put a government guarantee on them. Clearly, this will solve the 2008 crisis, but as always one needs to be looking forward. Bank crises are much less frequent in the 20th century than the 19th, but we need to be mindful of fixing the cause and not the symptom. Looked at in this narrative I marvel at how the future can be so like the past, but different enough to escape notice in real time. I suppose every war or financial crisis is like that.

Posted by Eric Falkenstein "

I do not know of a complicated model in any area of science that performs well in explanation and prediction

TO BE NOTED: From Falkenblog:

"Do Global Climate Models Contain Keynesian Macro Models?

In the 1970s there was a period when large scale, multi-equation macro models with hundreds of equations were used to forecast the economy. Those efforts were a dead end. Of course, the modelers themselves never admitted that, and those making macro models in the 1970s still make macro models today (see Macroeconomic Advisers LLC). Economics does not note approaches are failures, rather, it just stops doing them as the professors grow old and graduate students do not replace them.

Their demise was twofold. Chris Sims showed that a three instrument Vector Auto Regression (insanely simpler reduced form model) did just as well. His argument is applicable to any complicated model: when you have lots of interactions, every variable should be included in every equation, and so the model is underidentified because then there are more parameters than datapoints. Structural parameters with very different forecast implications are consistent with the same data, so you cannot forecast because you cannot identify the correct model. Secondly, no economic model, especially a large scale macro model, predicted that the socialist countries would severely underperform the capitalist countries, or that Africa would stagnate, or that the Asian tigers would flourish. Just about every major economic trend--computers, energy shortages, internet bubble, mortgage crisis--was seen only after they happened. Of course, they missed business cycles, and the secular decrease in inflation from 1980. If an economist were to tell me the average GDP for Paraguay in 2050, I imagine the more complex the model, the worse their forecast.

It is very easy to convince yourself that a complex model is correct, because you can always fine tune it to the data, and the pieces are never fully specified so outsiders can't judge how much it was fit to explain the data it is testing. For climate models it appears obvious there are more degrees of freedom in such a model than the two main datapoints driving them: the 0.75 degree Celsius increase in temperature and the 35% increase in atmospheric CO2 from 1900-2000. It is very suspicious that the models disagree a lot on specifics, like how much water vapor is expected over Antarctica, but generally agree on the temperature and CO2 implications. For a model with lots of interactions, one should see greater variability than what I have seen. I get the sense that to get funding one's work on a model has to be plausible, where 'plausibility' has already been decided.

Ronald Prin has a video on MIT's latest climate models, and he highlighted the economic aspects of the global climate model. Uh oh. He stated the climate models were as uncertain as the economic models. That's a high standard. Prinn admits to big uncertainties in climate models: clouds, which play a large role, are difficult to model. There are also uncertainties about emissions, and ocean-mixing, the churning of cooler and warmer waters, which can bring carbon buried on the ocean floor to the surface. He says there are 'hundreds' of these uncertainties. His solution is to look at the model under various changes in assumptions, generating hundreds of thousands of forecasts to estimate the probability of various amounts of climate change. As he states, “in the Monte Carlo sense, building up a set of forecasts on which we can put a measure of the odds of being correct or incorrect.”

But enumerating changes in assumptions and then looking at the effects is not an unbiased nor asymptotically consistent estimation process. Why should we believe the model is correct even if we knew the assumptions? You have, by Prin's estimate, 'hundreds' of uncertain parameters and processes, and forecasting over 100 years. Read Sokolove's outline, and you merely get graphics of 20 to 30 categories interacting in some massive feedback loop, and you have these for Land, Ocean, Urban area, and the Atmosphere, broken down by 20 regions, and then assumptions on technological changes (nuclear vs biofuel growth) as well as economic growth. Looking at Sokolov's description, it appears to be using a macro model within the super global climate model:

The current EPPA version specifically projects economic variables (GDP, energy use, sectoral output, consumption, etc.) ... Special provision is made for analysis of uncertainty in key human influences, such as the growth of population and economic activity, and the pace and direction of technical change.

Add to that the sociology-like graphs with big arrows pointing to A --> B --> A, enumerating sectoral employment growth in Modesto, and it looks a lot like those old, discredited large-scale macro models! Using a multi equation macro model in the Global Climate Model is like using a CDO squared formula to manage your portfolio.

I'm not saying there are much better models out there, just that these models are known to be useless for long term forecasting, and specifying every sector of the economy in such a manner seems like deliberate spurious precision given the known way such precision is treated by outsiders (eg, journalists) and the wealth of experience by insiders on how fruitless this approach is. These complex 'structural models' are like the early days of flight when people would strap on bird wings to fly because it emulated reality. Now we know that is a dead end, and use nothing like nature to fly. Specifying input-output relations by sector is a reasonable approach if it was not so thoroughly tried, tested, and found lousy.

Econometrician Arnold Zellner noted, after a lifetime of forecasting:

I do not know of a complicated model in any area of science that performs well in explanation and prediction and have challenged many audiences to give me examples. So far, I have not heard about a single one. ..it appears useful to start with a well understood, sophisticatedly simple model and check its performance empirically in explanation and prediction.

Instead of evaluating the models by how much they agree under various assumptions, I would want to see some step-forward, out-of-sample forecasting. That is, run the model through time T using data only from prior to T, and see how it does in period T+1. You do this through time and get a step foreward forecast. I haven't seen this. There are data on changes in CO2 and temperature using ice cores, but there is a causation problem: say the sun (or something exogenous) causes both warmth and CO2 rise. This will not allow you to estimate the effect of an exogenous CO2 rise on temperature unless you can identify an event with clear exogenous increases in CO2, as in a volcano. For example weight is correlated with height in humans: taller people tend to be heavier. But if you gain weight, that won't make you taller. The correlation we see over time in individuals, or cross sectionally, does not imply height is caused by changes in weight.

So many Global Warming proponents emphasize this is all about the facts and impugn the motives of global warming skeptics, yet the facts and theories underlying Global Warming scenarios are actually quite tenuous. There have been no important human-created trends or events that have been foreseen by a consensus of scientists. Adding that fact to the model implies we should ignore them.

Posted by Eric Falkenstein "