this was written by my friend Chris Martenson of chrismartenson.com who I think gives excellent and sober insight into the financial situation world wide.
He gave me permission to post this.
The bank stress tests, the results of which are to be announced later this week, have already failed. Why? Because the stress test is supposed to assess how a bank’s loan portfolio would fare under a variety of scenarios, but reality is already far worse than the worst-case scenario. The FDIC, Treasury, and Federal Reserve went ahead with their weak input assumptions to conduct the stress test, even though they knew that reality was outpacing their most dire scenario.
In other words, the stress test is a sham.
Think of it this way: Suppose someone asked you to assess how many cars out of a hundred would survive a drive from NYC to LA and back. You’d probably start with some simple questions about the age of each car, their past histories of mechanical difficulty, and the driving conditions. Now, suppose that after you’d made your best guess, it turned out that the cars had to drive the entire way on dirt roads, not highways, and in winter. You’d want to change your estimate, right?
Not the FDIC or the Treasury Department. They want to stick with their "stress test" results, even though conditions have worsened considerably since the stress tests were designed. In our analogy, banks are now driving on dirt roads in winter, but the Treasury and the FDIC, with great fanfare, will announce how those banks would perform if they were driving on highways in early summer.
My prediction: The stress tests will reveal that everything is mostly okay. For appearance's sake, a few banks will be shown to need a modest amount of capital, but this is only because the Treasury doesn’t think it can get away with simply proclaiming that everything is 100% good at every bank.
As always, the devil is in the details. Let’s peer in.
First, let’s discuss the purpose of the "stress tests." As the name implies, the idea here is to concoct a number of economic scenarios and apply them to each bank’s portfolio of assets to assess how that bank would perform under each circumstance.
Remember, to a bank, an "asset" is usually an outstanding loan of some sort, be it an auto loan, a mortgage, or a commercial loan. These days, we’d also have to include a variety of derivative products, such as the recently maligned Credit Default Swaps (CDS), Collateralized Mortgage Obligations (CMO), and the like. However, whether we are talking about derivatives or plain old loans, both behave the same: Under worsening economic conditions, each will suffer greater and greater losses.
If done correctly, a pretty good estimate of losses for each scenario can be developed. These losses can then be translated into additional capital needs for each bank and will need to be raised either from the private markets or another handout from future taxpayers.
There are two items that we have to trust in order to trust the results of the stress tests. The first is that the models used will provide accurate assessments, and the second is that the scenarios used will reflect actual stress-like conditions.
Hopefully, the models are nothing like the ones used by the ratings agencies, such as Standard & Poor’s or Moody’s, which rated many mortgage and derivative products as "AAA" when they were in fact pure junk. I have been unable to find any publicly available information on the interior workings of the stress-test models, beyond the next quote, drawn from a Federal Reserve document on the matter.
To guide estimation, the [banks] were provided with a range of indicative two‐year cumulative loss rates for each of the 12 loan categories for the baseline and more adverse scenarios. [Banks] were permitted to submit loss rates outside of the ranges, but were required to provide strong supporting evidence, especially if they fell below the range minimum.
Translation: Banks took the input assumptions and fed them into their internal models. If banks wanted to provide a lower estimate of loss, they were allowed to do so - but you and I cannot find out which banks did that, or how often, or the reasoning involved. Given the vast amount of so-called "level III assets" that these same banks are carrying on their books, assets which lack any sort of quantifiable market, I am extremely confident in stating that many banks will abuse this "out" and submit absurdly low loss rates.
It also bears noting that the "range of indicative two-year loss rates" are likely to be highly suspect and biased to the upside. I base this on the fact that we are breaking all historical records with respect to the pace and scope of housing price declines, and that other loan losses are increasing at a furious clip.
Every step of the way for the past two years, Federal Reserve economists have been entirely too optimistic, and I have little doubt that they remain biased to the upside here. That’s the nature of their job.
While there is no transparency in the models used or the loss ratios employed, the good news is that we can deduce everything we need to know about the stress tests by simply taking a close look at the parameters for the scenarios themselves.
As background for this examination, it bears noting that the stress test was announced to the nation by Sheila Bair of the FDIC on February 24, 2009, which tells us that the test was designed at some point in 2008. This is important because that’s when many of the baseline assumptions were set.
The stress test was designed with only two scenarios, a baseline and a "more adverse" scenario:
The supervisors developed an alternative "more adverse" scenario to reflect the possibility that the economy could turn out to be appreciably weaker than expected under the baseline outlook. By design, the path of the U.S. economy in this alternative more adverse scenario reflects a deeper and longer recession than in the baseline.
However, the more adverse alternative is not, and is not intended to be a "worst case" scenario. To be most useful, stress tests should reflect conditions that are severe but plausible.
I take exception with a stress test that aims to be "severe but plausible," as it betrays a belief that the future is predictable. The geniuses at LTCM got hammered by the "implausible," and every single economist at the Fed considered our current economic situation to be utterly "implausible" just 12 months ago.
But the really funny part here is that the conditions set for the "more adverse" scenario have already been either met or exceeded.
The following is a table of the stress-test assumptions used. I have placed green boxes around the "baseline" assumptions:
GDP, unemployment, and house prices are indeed the big drivers of bank loan losses, so these are sensible inputs. However, as we’ll soon see, even the "more adverse" scenario is already looking like an optimistic outcome.
GDP and the stress test
The baseline of the stress test assumes a -2.0% decline in GDP for 2009, followed by a 2.1% increase in 2010. My only question here is, "What were they smoking?"
Here’s the actual chart they released in February to reveal the GDP assumptions:
You might wonder how they could do this, given the fact that GDP declined by 6.3% in 2008Q4 and by 6.1% in 2009Q1. I have updated their chart to reflect this reality (below). Note that reality does not even fit on the stress-test chart. That blue line lurking below the chart frame on the lower left reflects the actual 2008Q4 and 2009Q1 results.
This means that reality is already running far below even the "more adverse" scenario, and the economy will have to absolutely skyrocket from here on out to avoid being worse than that. Also, you might note that once again, the assumption of a second-half recovery is baked into the estimates for both scenarios. There is no scenario that tests what might happen without a recovery, which is a distinct possibility, according to those who most accurately predicted this crisis in the first place.
Bottom line: The GDP assumptions for the stress test are wildly off the mark so far and are therefore not indicative of the actual stress being experienced by banks. Any stress test results that do not incorporate a more realistic assessment of economic growth/decline will be neither useful nor predictive.
Unemployment and the stress test
Below is a chart showing the unemployment assumptions that went into the stress tests. I have drawn black dotted lines on the chart at the maximum levels of unemployment for each scenario. As you can see, the baseline almost makes it to 9% by the second quarter of 2009 before backing off, while the "more adverse" case tops out at 10.5%.
Reality is already outpacing these scenarios. Below, I have added in a blue line for 20091Q unemployment, which was 8.1%.
Many private economists are now calling for unemployment to hit 10% in 2009, with some calling for rates of 11% to 12% in 2010. Are such rates possible? Certainly.
If we look at the current rate of unemployment on a 60-year chart, we can see that the current level is only the third highest in the series (so far). Given that this debt crisis has led to the shattering of so many economic records, both on the way up and now on the way down, it seems entirely defensible to suggest that unemployment could set a record of its own here.
Just for kicks, I took a chart of the most recent unemployment data and drew in some lines to indicate what each stress scenario implies, when compared to the most recent trend. All you need to do is decide for yourself which seems most likely.
For the baseline to occur, unemployment needs to immediately level off, while even the "more adverse" scenario would require a significant flattening of the current trajectory. The thing about unemployment, at least historically, is that it is a lagging indicator, meaning that it only levels off and turns up after everything else.
For the stress test scenarios to unfold, unemployment would have to behave in a way that it never has before.
Bottom line: The stress tests are using lowball, garden variety unemployment figures and reality has already outpaced even the "more adverse" scenario. This means that the stress test results will not be reliable indicators of actual banking losses or capital needs.
Housing and the stress test
The baseline stress test calls for a 14% decline in home prices this year, compared to 2008Q4, and a further decline of 4% in 2010. The "more adverse" case calls for a 22% decline this year and an additional 7% decline in 2010.
Even Goldman Sachs is calling for another 20% to 25% decline in home prices, putting them firmly in the "more adverse" camp. I say "even," because Wall Street investment houses are not famous for providing overly pessimistic calls. They usually err on the side of daffodils and rainbows.
I have a different view. I see this housing bubble as having started in 1998. And bubbles have a troubling tendency to burst all the way past their starting points.
Unfortunately, the Cash-Shiller Price Index does not go back to 1998, but it does go back to 2000, so I can make my case using 2000 as my starting point. Just keep in mind that 1998 was a lot lower than 2000, so I think that housing is going to fall a little further than the 2000 data would imply. So here goes:
The Case-Shiller Idex in 2008Q4 stood at a reading of 154.4. A 14% decline from there, as the baseline assumes, yields a reading of 132.8, a decline of slightly more than 21 points. The most recent reading in February of 2009 was 143.2, a decline of slightly more than 11 points. This means that in just the first two months of the year, fully one-half of the entire baseline decline in house prices has already been realized.
Oops.
Even compared to the "more adverse" case, a full third of that decline was logged in the first two months.
Double Oops.
To make this clearer, I placed the two stress-test declines on a chart of the Case-Shiller home prices going back to 2000. (Note: The Case-Shiller Home Price Index measures relative values of houses in 20 key markets. A reading of "100" in the year 2000 has been set as the baseline for the series in the chart below).
A few things jump out at me here. The first is that even the "more adverse" case does not fully unwind house prices to a level below their 2000 value, let alone their 1998 value when the bubble formed, although it does come close.
The second is that there’s nothing in the trajectory of the home prices that would lead me, as an analyst, to conclude that a 14% decline was a defensible position. A strong case would have to be made for why a significant leveling-out of the rate of price decline would occur. The current trend is far steeper than the baseline, and it has been for a year and a half.
I think a legitimate stress test should include the possibility of a decline to a value below the value seen in 2000. A relatively common feature of burst bubbles is that the asset in question goes below its starting point, which in this case is the reading of "100" set by the Case-Shiller Index.
A decline to 100 implies a drop of 35% from the 200Q4 reading, which is significantly deeper than the "more adverse" scenario in the stress test.
Bottom line: While house price declines are the most realistic of the stress test assumptions, they are still erring too far on the side of "optimistic" to be useful.
Conclusion
A "stress test," at least as far as I understand it from a scientific or engineering standpoint, is supposed to encompass a set of conditions beyond normal, or expected, values. The bank stress-test assumptions are already exceeded in each case by real-world conditions, and therefore will be neither illuminating nor predictive when they are released. Let’s all be thankful that the Federal Reserve and Treasury Department do not design bridges. If they did, they might "stress test" them by simulating an average load of traffic under average conditions and declare them perfectly safe.
I advise you to tune out what is certain to be an upbeat assessment of the condition of our banks, when the stress test results are finally released. My opinion is that the stress tests were specifically designed to be neither stressful nor revealing. Instead, they were designed to produce expected results for the purpose of instilling confidence in banks and in the crisis management team.
I would speculate that the stress test results will lead to a stock market rally, stronger bond prices, and falling gold prices. This has been a fairly reliable pattern after economic news is released, especially if the news is big enough and politically charged. The stress tests meet these criteria in spades.
Were I holding any stocks long that I wished to unload, I would use any such rally as the perfect vehicle to accomplish that task.
Because the stress test assumptions have already been exceeded by real-world conditions, we can readily assume that the actual condition of the banks is much worse than the stress test results.
Why do I spend so much time tearing these silly moments of economic history apart and then retelling them? I do this because it is all too easy to fall into the trap of believing these official distortions, which can cause confusion in one’s mind and paralysis of one’s plans.
By seeing these tricks and twists more clearly, I hope that you will be in a better position to ignore them, or even be amused by them, as you continue to work towards securing your own situation.