Monday, April 13, 2009

Roubini and the Stress Test Scenarios

by Bill McBride on 4/13/2009 09:08:00 AM

Professor Roubini writes: Stress Testing the Stress Test Scenarios: Actual Macro Data Are Already Worse than the More Adverse Scenario for 2009 in the Stress Tests. So the Stress Tests Fail the Basic Criterion of Reality Check Even Before They Are Concluded

[I]f you look at the actual data today macro data for Q1 on the three variables used in the stress tests – growth rate, unemployment rate, and home price depreciation – are already worse than those in FDIC baseline scenario for 2009 AND even worse than those for the more adverse stressed scenario for 2009. Thus, the stress test results are meaningless as actual data are already running worse than the worst case scenario.
I've noted before that the baseline case is no longer useful, and that the more adverse case is the new baseline. Roubini is taking this a step further and saying the more adverse case is also meaningless. I think this is premature - although I agree with Roubini that there is no real stress test.

First, Roubini is working from the annual stress test forecasts. The following table shows the quarterly data being used by the banks.

Quarterly Stress Test Click on graph for larger image in new window.

This table shows the quarterly GDP growth rate (annualized), unemployment rate, and house prices being used for the stress test scenarios.

House prices are based on the Case-Shiller Composite 10 Index with Dec 2008 = 100.

  • Unemployment

    For the unemployment rate, Roubini is correct. The unemployment rate was 8.1% in Q1 - above both the baseline (7.8%) and more adverse (7.9%) scenario rates.
    [B]ased on current and likely trends the unemployment rate will be – at best over 10% by year end – and more likely closer to 11% by year end (and average 9.8% for the year) 2009 data are already worse than the adverse scenario and will for sure be worse than the adverse scenario. But more importantly by year end 2009 the actual unemployment rate – even with a growth recovery in H2 – will be higher at 10.5% - than the average unemployment rate assumed by the FDIC in the adverse scenario for 2010, not 2009!
    In a recent note, S&P mentioned their stress test "assumes that an economic recovery does not begin until at least late 2010" and that "the unemployment rate rises to the mid-teens". That is a real stress test!

    From Bloomberg: Wall Street in Wells Fargo Moment as Euphoria Meets Stress Test
    “The bottom line is if the unemployment rate peaks at 10 percent these banks can make it through,” [Paul Miller, an analyst at FBR Capital Markets] said. “But if it peaks closer to 12 percent, nobody makes it. Or very few people make it.”
    So far the baseline case is meaningless, and for unemployment, the economy is tracking worse than the more adverse scenario.

  • GDP Growth Rate

    Roubini writes:
    A similar analysis suggests that the FDIC assumptions for GDP growth and home prices are already worse than the adverse scenario – let alone the baseline scenario – for Q1 of 2009. A first estimate of Q1 2009 GDP growth will be out only at the end of April 2009 but the current consensus is that the figure will be around -5% for the SAAR figure in Q1. ... the current consensus forecast for 2009 GDP growth is – at 3.2% - practically identical to the adverse scenario GDP growth for 2009; and most reputable research institutions are forecasting for 2009 a figure that is actually worse than the consensus scenario. Also, while the current consensus forecast for 2010 growth (2.0%) is practically identical to the baseline scenario for 20010 GDP growth (2.1%) a number of more accurate than consensus source are predicting a much weaker scenario for 2010: for example Goldman Sachs has a current forecast of 1.2% for 2010 GDP growth as opposed to the baseline scenario figure of 2.0%. So, in all likelihood even the current consensus forecasts is close – or worse – than the more adverse scenario while the baseline scenario is already out of the window both for Q1 and the year overall.
    Once again, Roubini is working off the annual stress test forecasts. The above table shows that a 5% annualized decline in real GDP is the baseline case.

    Earlier I compared the quarterly stress test forecasts with forecasts from Paul Kasriel at Northern Trust, and from Goldman Sachs (no link). See: Stress Test, Quarterly Forecasts for Unemployment and GDP

    Stress Test GDPThis graph compares the stress test forecasts for changes in real GDP with recent forecasts. Note: Kasriel has announced he is revising his Q1 GDP forecast upwards.

    An interesting note: the stress test scenario is using the advanced GDP release estimate for Q4 2008 of -3.8%, as opposed to the revised estimate of -6.2%.

    These bearish private forecasts are tracking slightly better than the more adverse scenario.

  • House Prices

    Roubini:
    [H]ome prices have been falling in recent months at a rate that is much higher than the 14% assumed in the FDIC baseline for 2009. They are also running currently at an annual rate that is higher than the 22% in the more adverse scenario of the FDIC; and even considering actual figures for the last few months – that show an accelerated rate of fall in homes prices between the spring of 2008 and the most recent data – home prices have been falling in the last few months at rates – average of y-o-y and m-o-m figures – of about 20% with an upward trend in the data. So the actual and trend figures are well above the baseline figure of 14% and closer to the 22% of more adverse scenario.
    As Roubini notes, we only have one month of data since the stress tests scenarios were released. The Case-Shiller index is released with almost a two month lag (January data was released near the end of March). Also, we have to be careful because there is a strong seasonal component to house prices.

    Case-Shiller Stress Test Comparison This graph compares the Case-Shiller Composite 10 index with the Stress Test scenarios from the Treasury (stress test data is estimated from quarterly forecasts).

    This is the first month and it is difficult to see the track on the graph. Here are the numbers:

    Case-Shiller Composite 10 Index, January: 158.04

    Stress Test Baseline Scenario, January: 159.69

    Stress Test More Adverse Scenario, January: 158.07

    It is only one month, but prices tracked the more adverse scenario in January.

    Roubini concludes in bold:
    Conclusion: Actual macro data for 2009 are already worse than the more adverse scenario in the stress tests. These are not stress tests but rather fudge tests
    I agree there is no real stress test, and the more adverse scenario is the real baseline. But I think it is premature to say that the more adverse scenario is meaningless.

  • Last 10 Posts