Coverage and Collateral

I thought that I must be just woefully old fashioned. 

In my mind the real reason for the financial crisis was that bankers lost sight of what it takes to operating a lending business. 

There are really only two simple factors that MUST be the first level of screen of borrowers:

1.  Coverage

2.  Collateral

And banks stopped looking at both.  No surprise that their loan books are going sour.  There is no theory on earth that will change those two fundamentals of lending. 

The amount of coverage, which means the amount of income available to make the loan payments, is the primary factor in creditworthiness.  Someone must have the ability to make the loan payments. 

The amount of collateral, which means the assets that the lender can take to offset any loan loss upon failure to repay, is a risk management technique that insulates the lender from “expected” losses. 

Thinking has changed over the last 10 – 15  years with the idea that there was no need for collateral, instead the lender could securitize the loan, atomize the risk, thereby spreading the specific risk to many, many parties, thereby making it inconsequential to each party.  Instead of collateral, the borrower would be charged for the cost of that securitization process. 

Funny thing about accounting.  If the lender does something very conservative (in terms of current standards) and requires collateral that would take up the first layer of loss then there will be no impact on P&L of this prudence. 

If the lender does not require collateral, then this charge that the borrower pays will be reported as profits!  The Banks has taken on more risk and therefore can show more profit! 

EXCEPT, in the year(s) when the losses hit! 

What this shows is that there is a HUGE problem with how accounting systems treat risks that have a frequency that is longer than the accounting period!  In all cases of such risks, the accounting system allows this up and down accounting.  Profits are recorded for all periods except when the loss actually hits.  This account treatment actually STRONGLY ENCOURAGES taking on risks with a longer frequency. 

What I mean by longer frequency risks, is risks that expect to show a loss, say once every 5 years.  These risks will all show profits in four years and a loss in the others.  Let’s say that the loss every 5 years is expected to be 10% of the loan, then the charge might be 3% per year in place of collateral.  So the banks collect the 3% and show results of 3%, 3%, 3%, 3%, (7%).  The bank pays out bonuses of about 50% of gains, so they pay 1.5%, 1.5%, 1.5%, 1.5%, 0.  The net result to the bank is 1.5%, 1.5%, 1.5%, 1.5%, (7%) for a cumulative result of (1%).  And that is when everything goes exactly as planned! 

Who is looking out for the shareholders here?  Clearly the deck is stacked very well in favor of the employees! 

What it took to make this look o.k. was an assumption of independence for the loans.  If the losses are atomized and spread around eliminating specific risk, then there would be a small amount of these losses every year, the negative net result that is shown above would NOT happen because every year, the losses would be netted against the gains and the cumulative result would be positive. 

Note however, that twice above it says that the SPECIFIC risk is eliminated.  That leaves the systematic risk.  And the systematic risk has exactly the characteristic shown by the example above.  Systematic risk is the underlying correlation of the loans in an adverse economy. 

So at the very least, collateral should be resurected and required to the tune of the systematic losses. 

Coverage… well that seems so obvious it doed not need discussion.  But if you need some, try this.

Advertisements
Explore posts in the same categories: Compensation, Correlation, Debt, Diversification, Financial Crisis, Leverage, Revenues, Sub prime

Tags:

You can comment below, or link to this permanent URL from your own site.

2 Comments on “Coverage and Collateral”

  1. riskviews Says:

    Final sentence is a really interesting point. Some truth to it, but . . .

    Some of the excess liquidity was created from the housing price bubble. Not sure how much, but in every bubble there comes a point where the bubble feeds itself.

    Additions to value of assets creates additional value that can be used as additional collateral for additional borrowing for more investing in the asset class which creates additional demand, which causes additions to value.

    Other liquidity increases came from big surges in the amount of leverage of banks. I hear numbers north of 30 to 1 for a couple. And I am not sure how that treats the leverage built into many derivative contracts. It might just be ignoring that.

    So there was certainly some excess liquidity that was not a part of the bubble behavior, but I am not at all sure how much.

  2. nnja Says:

    A few points to comment on here. Excess spread on a subprime loan was not a replacement for collateral, it was to reflect the additional risk of the borrower. This is evidenced by the fact that a 2006 vintage prime borrower of the same property would not be charged an extra 300 bps.

    The other problem is a larger concern about default or loss incidence. I don’t think that there was a 10% chance of a borrower defaulting for each year of the decade, so that it was likely that one year would be a bad year. Instead, during times of relative sanity, there was, say, a 2% chance of default, and by 2006 and 2007, say, a 80% chance of default. Failing to recognize that distributions remain homogeneous for only a short time is one of the biggest challenges to risk management. 2006 was not a bad year, caused by fate, or Tyche, it was a year where a lot of bad decisions were made, with relatively predictable consequences.

    On the one hand, a senior secured loan on just about anything should never have a loss approaching (or even exceeding!) 100% – this was a failure to be sufficiently collateralized (e.g. not recognizing the possibility of a bubble, allowing 100%+ financing, etc.). But on the other hand, you can’t make a good loan to a bad credit. These loans were never designed to be paid down via amortization, they were expected to be refinanced (the greater fool theory for creditors). So instead of 30% of subprime borrowers defaulting, you get 80%. Making a no doc loan with ridiculous DTI ratios led to the high default rates, and therefore the failure of the product, but I think that the failure of the system came from the high loss severity. Making loans to people with bad credit has always been a risky venture, but using real estate as collateral has generally reduced risks substantially. Not this time.

    The high loss severities came from the fact that housing prices in many areas were in an unsustainable bubble. But the bubble only occurred because banks and mortgage lenders would make a loan for just about any amount to just about anyone. So if the industry hadn’t walked off a cliff in terms of lending to bad credits, I don’t know if there would have been enough money entering the system to create a housing bubble in the first place, which would have maintained the value of the collateral. If this were to hold, then it means that better underwriting of the borrowers would have inadvertently led to better underwriting of the collateral.

    Of course, the excess liquidity in the system had to go somewhere, so maybe even if the housing bubble was unavoidable, the financial crisis was not.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s


%d bloggers like this: