Badly Shaved Monkey
Anti-homeopathy illuminati member
- Joined
- Feb 5, 2004
- Messages
- 5,363
I have been trying to make some sense of the financial crisis we all seem to be caught up in, but I have been struck by what seems to be a contrast between the headlines telling us we have struck an iceberg and are sinking and what seem to be the data underpinning those headlines.
Typical headlines and reports concentrate on the "millions" of Americans defaulting on their mortgages and the huge rises in foreclosures.
But there are always "millions" of Americans defaulting on their mortgages, America is a big place and a minority of mortgages defaults is a permanent feature of economic life.
http://www.mbaa.org/NewsandMedia/PressCenter/64769.htm
It seems to me there has been a seriously misrepresentative use of relative changes to small absolute risks comparable to the misuse of "relative risk" data in medical scares that Ben Goldacre highlights in his book.
To draw some numbers from that MBAA page. Foreclosures have risen from about 1.45% to about 2.75%. That is a doubling, but it is an absolute rise of only 1.35%. Furthermore, that rise is almost by definition likely to be biased towards the lower end of the property-value scale, so the increased proportion of mortgage lending that is at risk is <1.35%.
Or, from another source:
"The proportion of homeowners delinquent on their mortgages but not yet in foreclosure reached 5.82 percent during the quarter, the highest level since 1985. That was up from 4.95 percent during the fourth quarter of 2006."
In the subprime area, there is obviously a specific problem (MBAA again),
"The delinquency rate decreased 12 basis points for subprime loans to 18.67 percent"
But I have seen estimates that subprime lending constitutes about 5% of the total, so even at a level of nearly 20% delinquencies, that is still only 1% of total lending, which aligns with the "<1.35%" I inferred above.
To add a little UK perspective on a related theme, we are given reports of the various House Price Indices and most often the news reports those produced by the Halifax and Nationwide Building Societies. These show double-digit price drops over 1yr. But, the less well-reported index produced by the Land Registry, that is based on actual price paid shows a much smaller rate of decline. Currently 4.6%. What is the difference? The lenders' indices are based on valuations made at the time that the loan is agreed. Clearly there are many factors at play there that make this different from price actually paid. The media seem to choose the scarier but less reliable indices.
Obviously there is a problem, but is the real problem more one of perception that the underlying economic situation: the headline bank-crashes reflect that minority of the market that geared itself tightly to the worst performing loans. Are we not now in a state where the fears of crisis are making that crisis a self-fulfilling prophecy more than the actual size of the necessary write-down in equity values?
In summary, if we could just accept that we're all 1-2% poorer than we were a year ago, would the current panic evaporate?
Typical headlines and reports concentrate on the "millions" of Americans defaulting on their mortgages and the huge rises in foreclosures.
But there are always "millions" of Americans defaulting on their mortgages, America is a big place and a minority of mortgages defaults is a permanent feature of economic life.
http://www.mbaa.org/NewsandMedia/PressCenter/64769.htm
It seems to me there has been a seriously misrepresentative use of relative changes to small absolute risks comparable to the misuse of "relative risk" data in medical scares that Ben Goldacre highlights in his book.
To draw some numbers from that MBAA page. Foreclosures have risen from about 1.45% to about 2.75%. That is a doubling, but it is an absolute rise of only 1.35%. Furthermore, that rise is almost by definition likely to be biased towards the lower end of the property-value scale, so the increased proportion of mortgage lending that is at risk is <1.35%.
Or, from another source:
"The proportion of homeowners delinquent on their mortgages but not yet in foreclosure reached 5.82 percent during the quarter, the highest level since 1985. That was up from 4.95 percent during the fourth quarter of 2006."
In the subprime area, there is obviously a specific problem (MBAA again),
"The delinquency rate decreased 12 basis points for subprime loans to 18.67 percent"
But I have seen estimates that subprime lending constitutes about 5% of the total, so even at a level of nearly 20% delinquencies, that is still only 1% of total lending, which aligns with the "<1.35%" I inferred above.
To add a little UK perspective on a related theme, we are given reports of the various House Price Indices and most often the news reports those produced by the Halifax and Nationwide Building Societies. These show double-digit price drops over 1yr. But, the less well-reported index produced by the Land Registry, that is based on actual price paid shows a much smaller rate of decline. Currently 4.6%. What is the difference? The lenders' indices are based on valuations made at the time that the loan is agreed. Clearly there are many factors at play there that make this different from price actually paid. The media seem to choose the scarier but less reliable indices.
Obviously there is a problem, but is the real problem more one of perception that the underlying economic situation: the headline bank-crashes reflect that minority of the market that geared itself tightly to the worst performing loans. Are we not now in a state where the fears of crisis are making that crisis a self-fulfilling prophecy more than the actual size of the necessary write-down in equity values?
In summary, if we could just accept that we're all 1-2% poorer than we were a year ago, would the current panic evaporate?