Thursday, November 29, 2007

A Tribute to Minsky, Part I

Hyman Minsky is someone who I have mentioned a few times on this blog, yet I've noticed that I have not ever shared my own interpretations on his work. So in Part 1 of my tribute to Hyman Minsky, I will be going over "The Financial Instability Hypothesis", and what I have built on his idea.

The first thing to understand is the idea of debt-income relationships. This term refers to the idea that a borrower's debt relies on the borrower's income to get paid back. But I think underlying this is an even more general and important relationship: price-value. Below is a chart I came up with.


For now, we will focus just on debt and investment ( I threw the idea of consumption in there just because I know it too plays a part and to remind myself about this later) Debt is a relationship whose value is derived from the borrower's income, and an investment is similarly has its value derived from the underlying cash flow. Now there are some differences between these two. For example, with debt the upside is capped. The borrower's income can triple, but he will most certainly not start paying any increased amount to his bank. But in an investment, any increase in cash flow accrues to the owner. So, we see that:

With Debt, the maximum upside is known and quantifiable. All that exists is downside.
With Investment, the range of values is infinite, meaning there is no limit to your potential gain.

Nassim Talib described this phenomenon as negative and positive black swans. (debt and investment, respectively. See Post) Otherwise, I think you can see that the relationships are fairly similar.

Now the essence of Minsky's work revolved around the idea that debt-income could be characterized into 3 broad categories- hedge, speculative, and ponzi. In hedge financing, the borrower's income can cover both his interest payments and his eventual principal payments. In speculative financing, the borrower's income can cover his interest, but he must either refinance or sell assets to pay the principal. Finally, in Ponzi financing the borrower's income can not cover either the principal or the interest payments. Some people might already be thinking something along the lines of:

"Hey, not even able to afford their interest payments? Isn't that what is causing this mortgage crisis right now, with loans being made to people who could not afford it?"

With the answer being yes. But what is interesting for value investors is that these three classifications apply to the broader price-value spectrum as well. Let's take common stocks for example. We have a purchase price, which is the market capitalization we are paying, and we have the value, which is the company's cash flow. If the company's cash flow can cover both our interest (cost of capital) and the principal (purchase price), then we have a hedge financing relationship. Worded differently, if we can discount the cash flow(cost of capital) and get a value more than the principal (purchase price), then we are hedge financing. Sound familiar? It should, because this is the dictum that value investors live by: A common stock is worth the discounted value of its future cash flow. Our goal is to buy something for less than that.

So let's move down the chain a bit. A speculative arrangement in stocks would mean that the cash flow can cover the cost of capital, but it can not cover the purchase price. And with Ponzi, the cash flow can not even cover the cost of capital. Now in the world of stocks, this environment has usually been referred to as a bubble. Think of the tech bubble as an example. Many companies were trading at earnings yields of 2% or less, while treasury rates were north of 6%. So, an investor in common stocks at that time could not cover his cost of capital out of cash flow.

This analogy has helped me describe the current mess we're in as a true Credit Bubble, or as Minsky would say a Ponzi. Loans were made on a large scale to people who could never afford to make the true cost of interest out of their incomes, yet adjustable rate mortgages helped mask this for some time.

Now some people have tried to claim that the problem will not be so bad because these loans are secured by homes, which have real value. But again, we can apply the all too familiar concept from value investing to show this is not the case. An asset is worth the discounted sum of its future cash flows. We've seen how this is the case for common stocks. A home is the same thing, although with the value being derived from the property's rent. And using current rents, homes as an asset are also in a Ponzi relationship, not being able to cover the cost of capital.

Now it is important to also realize that just because current cash flow does not cover interest doesn't mean it is necessarily a bad deal. I'll use another stock example to make the point. It is perfectly reasonable for me as an investor to invest in Coke at lets say a 4% earnings yield when the cost of capital is 6%, and for it to still be a value. This is because I expect Coke's cash flow to continue to grow into the future. So although currently my cash flow can not cover interest, in the near future it will. If I am correct and Coke's Net Present Value is greater than the price I am paying for it, then it is in actuality still a hedge financing arrangement.

So is there a similar hope for mortgage debt and housing? Well, mortgage debt is based off borrower's income. And housing is derived from rent, but rent is also a function of income. So for these loans to work out, we would need to see incomes rise in the future. And the recent trend is not encouraging. Below is a graph of REAL household income from 1967 to 2005. Nominal would be more appropriate, but I could not find a chart for it, and real incomes is suitable. Since 2000, household incomes have actually declined. An important factor in this has likely been globalization, which I have discussed in the "Tectonic Shift" series. With the competition of the entire world's labor market, it is difficult to assume incomes will be able to rise considerably in the near future.




So for now, that is the dilemma we face. In part two, I will try to expand on this concept and describe some of the possible effects.

No comments: