Friday, October 31, 2008

Save GM's workers, but not GM

I've just listened to a discussion on KQED about whether the US government should bail out the auto industry. The reason the solution here is not obvious is that everybody assumes that to save the job of someone who works for GM, you have to save GM. I suspect that's not true.

Toyota makes lots of cars in the United States. I suspect the majority of Toyota's domestic sales are domestically built. It saves them shipping, and avoids exposure to foreign exchange risk. The overall point is that there is no problem with American workers making cars that sell. The problem is that some American workers are making the wrong cars, and that is the fault of their engineering and management.

The usual way for these things to work out is that GM would go bankrupt and it's assets would be sold. Toyota would hire some of those workers, and buy some of those assets on the cheap, because they'd know that with GM gone there would be a reduction in the supply of cars and so they could make more money by increasing their own production.

Crucially, the assets (e.g. assembly plants) would be sold for a fraction of their original cost because Toyota would have to adapt them to Toyota's production style. This is very important -- GM's production style isn't profitable, so those assets have to be changed to become profitable.

Also, crucially, the former GM workers hired by Toyota would have to be retrained. This takes a lot of time for Toyota. The rehiring process also tends to weed out at least some of the poorly producing workers.

All this adaptation of physical capital and retraining of workers is all good stuff. So what's so bad about bankrupcy?

GM's shareholders lose a lot of money. Actually, GM's shareholders lost a lot of money a long time ago. GM's total market cap isn't very large, and hasn't been for a long time, because GM has had bad management for a long time. I claim that this problem is not terrible and does not warrant intervention from the government.

GM's workers and suppliers have no business for a long time. This is the real problem, and this is something that the government can help with.

So, here is what I suggest. I suggest that because GM, Ford, and Chrysler are "too big to fail", the government should intervene before they shut their doors, and negotiate an orderly transfer of their assets to companies that can make those assets perform. I suggest that Toyota and Honda and Mercedes and so forth would be willing to purchase much of the assets of the big three for a fraction of their orginal price, and would be willing to take government inducements to keep the factories open and the workers employed while they are restructured and retrained.

The result will be that some assembly people will be put out of work, because not all the Big Three's plants will be purchased. The corporate management of the big three will go unemployed, which is fine, as they are responsible for the current distress in those companies. There will be differences between the financial commitments that GM has made to its workers, and those that Toyota makes its own, and the differences between those commitments will have to be worked into the sale price from GM to Toyota.

We need to clarify the idea of "too big to fail". Our commitment is to our people, not to our companies.

Wednesday, October 29, 2008

The Biggest Haul Ever

The $700 billion bailout is about the same size as the dollar cash money supply in North America (that is, excluding dollars held as reserve currency in foreign sovereign banks).

The robbers in the Securitas Depot Robbery of 2006 made off with $92 million. This is the largest robbery in the western world, surpassed only but hugely by the $1 billion heist of cash from the Central Bank of Iraq in 2003, which was curiously similar to current events in that it appears to have been made possible by the actions of the Republicans in the U.S. government. In that case, it was the initiation of bombing of Bagdad. The current mess traces back to the Gramm-Leach act of 1999 which repealed the Glass-Steagall Act of 1933, which was intended to prevent banks from engaging in risky investment practices such as those that had caused the banking collapse in early 1933.

But neither of those robberies is anything like the size of the bailout. Imagine that some group managed to knock over every single bank, not just in one city, but in all of them. Not just the banks, but the convenience stores, groceries, businesses. Not just those but swiped the cash from every house and every wallet of every person in the United States. What a haul! It's positively Grinchian in scope.

Now imagine that those responsible are being paid a salary by the federal government. And that the folks they are turning the money over to are being paid extra to figure out what to do with it!

I've been mugged!

Wednesday, October 01, 2008

Bail out Main Street, not Wall Street

The $700 billion Paulson plan is too expensive and fixes the wrong end of the problem.  A better solution is obvious: bail out the mortgages that are going into default, rather than bailing out the banks that are capitalized by derivatives of those mortgages.

Bailing out the mortgages themselves is much cheaper (around $200 billion), directly eases the problems of millions of homeowners, and should fix the bank's capitalization problems.

Now, if it doesn't fix the capitalization problems, that means there is a lot more bad debt floating around than just these subprime mortgages.  And if this other debt is the problem, let's get that problem front and center so we can deal with it.

Some background:  The subprime crisis is essentially this:
  • Millions of people in the U.S. bought houses more expensive than they could afford, on the assumption that the rise in the price of the house would help them pay for the house.
  • Those people accepted mortgages with initally low rates, that then increase after a period of time.  They figured that before the rates increased, they would either get a job that paid better, or refinance the house based on a higher assessed value which would give them a larger equity stake and therefore make it possible to get a loan with a lower interest rate.
  • Housing prices are not rising, and those people are unable to refinance and don't have a better job.  When the interest rate rises, they can't pay the increased bill and fall behind on their mortgage payments.
After 10 years of excessive spending on houses, and excessive building of expensive houses, the US has too many expensive houses.  The house price collapse will not be resolved until the US economy has grown enough folks with enough income to afford those houses.  My guess is this will take about a decade.

In the interim, we cannot allow those extra expensive houses to go unoccupied.  Unoccupied houses fall apart, and so they lose value quickly.  This loss will be borne by the large institutions in our economy and will be a net drag on our net asset growth.  Under the Paulson plan, the government will buy derivatives of these mortgages and these losses will pass to the taxpayer.

So, the houses must remain occupied, and we do not have enough people who can afford them to occupy them.  Until we grow those folks, the best people to occupy those houses is the people who are living in them now.  I have a plan to keep them there, detailed below.

The financial crisis is an extension of the subprime crisis:
  • Banks have been converting bundles of mortgages into mortgage-backed equities, and selling those.  Many institutions now own these assets.
  • The value of those equities depends on the number of mortgages within them that default.
  • The value was initially determined by making assumptions about the usual number of mortgage defaults.  Side note: some mortgage-backed equities are guaranteed, so that the risk of many defaults is seperated out into another security.  That does not change the problem, only who owns it.
  • Because the subprime crisis is surprising, nobody is comfortable estimating the number of mortgage defaults any more, and so the value of these equities is not well known.  Their value is plummeting for two reasons:
    • More risk means less value.
    • Banks are trying to sell these MBEs, but nobody is buying.
  • Banks are required to have a dollar of capital for every 12 dollars they loan out.  When the value of their assets drops, they have to sell some of their loans (packaged as mortgage-backed equities) to generate more capital.
  • So, selling MBEs causes the value of the MBEs still held to drop, which forces the bank to sell more MBEs.  The feedback loop is driving the capitalization of the banks below the legal requirements, which causes the banks to fail.
Paulson's plan is to have a U.S. government agency buy $700 billion of mortgage backed assets at some "fair" price.  This price will set a floor on the valuations of the assets owned by banks, so that they don't have to sell so many MBEs, and it will give them the cash they need to improve their asset positions to the point where they can loan money again.

The Paulson plan will fix the Wall Street problem, but it does not address the Main Street problem, that is, the mortgage defaults, except that in avoiding a recession it will reduce the number of mortgage defaults from people losing their jobs.

The alternative is to fix the Main Street problem directly.  If the government guarantees that only the expected number of mortgage defaults happen, then the MBE will have known values, which should also put a floor on the valuation of the bank's held assets and similarly ward off the credit crunch.

Here is how the government makes that guarantee.  Let's say that Harry has an adjustable-rate mortgage which just increased its rate, and he can't afford it.  His original 10% equity in the house is down to 5%.  Ordinarily he would go into default, and at some point the bank would reposess the house and evict him, and then sell the house.  We don't want that to happen because either the house will go vacant for a long period of time, during which it will deteriorate and lose value, or the bank will sell it for a song and lose a great deal of money, causing the bank to go bankrupt.

Instead, the government steps in and offers Harry the following deal:
  • The government buys 25% of the house from Harry.
  • Harry now refinances his 75% of the house.
    • He now has 6.66% equity in his smaller share of the house.
    • His loan amount is 74% of what it was, so his payments are proportionally smaller.
  • At some point in the next 10 years, presumably after the housing market has worked through its excess inventory, Harry gets a one-year notice from the government that he needs to either buy back the government's 25% or sell the house.  He can sell at any earlier time if he likes.
So, how much cash would the government have to invest with this plan?  Suppose the number of subprime mortgages is 9 million, and 25% of those are in serious delinquincy.  Suppose the average value of those houses was $320,000.  For the government to purchase a 25% stake in all of those would cost $180 billion.  And note that this is an investment.  After ten years or so, we should expect to see most of that money back, perhaps with some gain.

Why is my plan cheaper than Paulson's plan?  The primary difference is that my plan keeps people in their houses by purchasing a portion of the house's value.  Paulson's plan requires the government to buy the full value of the mortgage.  What Paulson's plan does not detail out is that the government will be stuck with the loss when the owners default on the mortgage and then the house is repossessed and sold for a loss.