Dr. John Rutledge Blog, September 30, 2008
My friend Sonia Arrison, the technology wizard at Pacific Research Institute, asked me to give her my current thinking about how to analyze the current credit crunch as a cascading network failure, a subject I wrote about in Chapters 6 and 7 of my new book.
What? You don’t have a copy of Lessons from a Road Warrior yet? Quelle tragedie! I recommend getting one for every member of your family. In fact, send one to your Congressman too. He obviously needs to read something! You can get one at our website by clicking on the title above. It will be up on Amazon in a week or so.
The current crisis is a perfect example of a massive cascading information network failure.
Markets are so efficient in processing information (using prices to carry information, as Hayek wrote 70 years ago) that when the network fails, and we move to nonprice rationing (credit crunch) the resulting drop in efficiency causes a disruption in both the capital markets and the GDP economy.
In this case, as I have been screaming loudly since the White House passed their useless stimulus program, one big mistake was viewing the issue as a typical macroeconomic problem where people don’t spend enough money so the cure is to boost spending. This crisis is a capital market failure, not a weakness in spending. That’s why we have been having bank failures but no recession (positive GDP.) At least so far.
The capital market problem is that bond market investors lost faith in their ability to predict the cash flows they would receive as owners of the asset-backed securities. With no visibility they could not estimate value. They just backed away from the table.
Last summer, the day after Blackstone’s IPO. The very next day the financing failed for a large private equity deal. That summer all the headlines were about the banks’ owning of $300 billion of toxic leveraged debt. Next, in August, it spread to mortgage-backed securities.
The asset-backed securities market has a fatal flaw. The securities are created by using historical correlation coefficients to slice a pool of mortgage’s cash flow into pieces with different risk characteristics. The work is done with the Black-Scholes theorem which, like modern portfolio theory and CAPM, assumes that the market for the securities IS ALWAYS IN EQUILIBRIUM, i.e., that there is never a situation of nonprice clearing, never a network failure. But network failures happen, like the October 1987 portfolio insurance crash, the Russian bond default, the S&L crisis and RTC, and LTCM. When correlations change (which the MPT guys have the balls to call six sigma events), as they always do, investors lose all ability to “see” the cash flows they will receive and the market collapses.
Network theory tells us that every network will eventually fail. The robustness of a network, i.e., how frequently, how deeply, and for how long the failure will last is determined by the system’s “architecture”. i.e. by the pattern of connections between the nodes. Networks like our capital market, where a few nodes (the Fed, Treasury, FNMA, Wall Street banks) are connected to a huge number of nodes (smaller banks, companies, mortgage borrowers) is called a “scale free” network. It’s frequency distribution is a power law. When one of the “supernodes” fails everything goes down and the network fails.
The Fed, for their part, has made this worse through their policy of targeting the Fed funds rate. Fed funds targeting, like Black-Scholes, assumes the market is in price-equilibrium. In markets where there is nonprice rationing–such as the current credit crisis–the posted fed funds rate is not a true measure of the cost of funds, just as the monthly rent is not a measure of the value of an apartment during rent control periods.
I wrote a few days ago that the Fed is out of touch. A month ago they were worrying about inflation risks. The Federal Open Market Committee members are equilibrium thinkers. They think a low fed funds rate means there is plentiful credit (it does not.) They focus on the wrong prices, thinking their job is to control the prices of currently produced goods and services (the CPI, PPI, and GDP deflator). It is not. Their job is to control the price level for the massive stock of tangible wealth, made up of land, houses, used cars, capital goods and other hard assets. Those prices are deflating big time. Get a life.
Since March, they have been providing hail mary liquidity lines to troubled institutions like Bear Stearns, Lehman, Fannie Mae and AIG. But they have a sterilization policy that triggers a simulaneously sale of Tbills (shrink bank reserves) by an equal amount each time.
As a result, bank reserves have not increased in any meaningful way for the past year. (Total reserves on Sept. 10, just two weeks ago, were actually lower than they were a year earlier.) The Fed has been squeezing on a balloon, taking reserves away from a healthy bank to give to a sick one. That’s why banks have been falling like dominoes, one by one. That’s why the Fed’s balance sheet of Treasuries has shrunk from about $800 billion to about $500 billion over the past year. Until they significantly increase reserves this show will continue. (The only exception is last week, when we finally saw a huge jump in reserves and the monetary base. Hope they have turned over a new leaf but I am skeptical.)
Last time we had a capital market blackout (2000-2004) I wrote an op-ed in the WSJ saying that we should change the regulatory structure to push power down to local regulators (to break up the supernodes). I was attacked by the Comptroller of the Currency for my temerity.
In China, I am working with senior leaders on the development of their new capital markets, trying to get them to build markets where failures would be less frequent, more shallow, and of shorter duration.
Final point. Yesterday I think we had a second network failure–a cascading failure of the political network. That’s why the vote failed. The turbulence I wrote about in Chapter 6 has generated a huge class war between those of us who are fortunate enough to play with capital or technology and those of us who still hold a wrench. We see it in the campaign speeches. We see it on the television. And we saw it in the rejection of the rescue plan as a rich guy bailout.
I am very concerned that we have not seen the worst of this.
The Credit Crunch is a Network Failure
John Rutledge
Dr. John Rutledge Blog, September 30, 2008
My friend Sonia Arrison, the technology wizard at Pacific Research Institute, asked me to give her my current thinking about how to analyze the current credit crunch as a cascading network failure, a subject I wrote about in Chapters 6 and 7 of my new book.
What? You don’t have a copy of Lessons from a Road Warrior yet? Quelle tragedie! I recommend getting one for every member of your family. In fact, send one to your Congressman too. He obviously needs to read something! You can get one at our website by clicking on the title above. It will be up on Amazon in a week or so.
The current crisis is a perfect example of a massive cascading information network failure.
Markets are so efficient in processing information (using prices to carry information, as Hayek wrote 70 years ago) that when the network fails, and we move to nonprice rationing (credit crunch) the resulting drop in efficiency causes a disruption in both the capital markets and the GDP economy.
In this case, as I have been screaming loudly since the White House passed their useless stimulus program, one big mistake was viewing the issue as a typical macroeconomic problem where people don’t spend enough money so the cure is to boost spending. This crisis is a capital market failure, not a weakness in spending. That’s why we have been having bank failures but no recession (positive GDP.) At least so far.
The capital market problem is that bond market investors lost faith in their ability to predict the cash flows they would receive as owners of the asset-backed securities. With no visibility they could not estimate value. They just backed away from the table.
Last summer, the day after Blackstone’s IPO. The very next day the financing failed for a large private equity deal. That summer all the headlines were about the banks’ owning of $300 billion of toxic leveraged debt. Next, in August, it spread to mortgage-backed securities.
The asset-backed securities market has a fatal flaw. The securities are created by using historical correlation coefficients to slice a pool of mortgage’s cash flow into pieces with different risk characteristics. The work is done with the Black-Scholes theorem which, like modern portfolio theory and CAPM, assumes that the market for the securities IS ALWAYS IN EQUILIBRIUM, i.e., that there is never a situation of nonprice clearing, never a network failure. But network failures happen, like the October 1987 portfolio insurance crash, the Russian bond default, the S&L crisis and RTC, and LTCM. When correlations change (which the MPT guys have the balls to call six sigma events), as they always do, investors lose all ability to “see” the cash flows they will receive and the market collapses.
Network theory tells us that every network will eventually fail. The robustness of a network, i.e., how frequently, how deeply, and for how long the failure will last is determined by the system’s “architecture”. i.e. by the pattern of connections between the nodes. Networks like our capital market, where a few nodes (the Fed, Treasury, FNMA, Wall Street banks) are connected to a huge number of nodes (smaller banks, companies, mortgage borrowers) is called a “scale free” network. It’s frequency distribution is a power law. When one of the “supernodes” fails everything goes down and the network fails.
The Fed, for their part, has made this worse through their policy of targeting the Fed funds rate. Fed funds targeting, like Black-Scholes, assumes the market is in price-equilibrium. In markets where there is nonprice rationing–such as the current credit crisis–the posted fed funds rate is not a true measure of the cost of funds, just as the monthly rent is not a measure of the value of an apartment during rent control periods.
I wrote a few days ago that the Fed is out of touch. A month ago they were worrying about inflation risks. The Federal Open Market Committee members are equilibrium thinkers. They think a low fed funds rate means there is plentiful credit (it does not.) They focus on the wrong prices, thinking their job is to control the prices of currently produced goods and services (the CPI, PPI, and GDP deflator). It is not. Their job is to control the price level for the massive stock of tangible wealth, made up of land, houses, used cars, capital goods and other hard assets. Those prices are deflating big time. Get a life.
Since March, they have been providing hail mary liquidity lines to troubled institutions like Bear Stearns, Lehman, Fannie Mae and AIG. But they have a sterilization policy that triggers a simulaneously sale of Tbills (shrink bank reserves) by an equal amount each time.
As a result, bank reserves have not increased in any meaningful way for the past year. (Total reserves on Sept. 10, just two weeks ago, were actually lower than they were a year earlier.) The Fed has been squeezing on a balloon, taking reserves away from a healthy bank to give to a sick one. That’s why banks have been falling like dominoes, one by one. That’s why the Fed’s balance sheet of Treasuries has shrunk from about $800 billion to about $500 billion over the past year. Until they significantly increase reserves this show will continue. (The only exception is last week, when we finally saw a huge jump in reserves and the monetary base. Hope they have turned over a new leaf but I am skeptical.)
Last time we had a capital market blackout (2000-2004) I wrote an op-ed in the WSJ saying that we should change the regulatory structure to push power down to local regulators (to break up the supernodes). I was attacked by the Comptroller of the Currency for my temerity.
In China, I am working with senior leaders on the development of their new capital markets, trying to get them to build markets where failures would be less frequent, more shallow, and of shorter duration.
Final point. Yesterday I think we had a second network failure–a cascading failure of the political network. That’s why the vote failed. The turbulence I wrote about in Chapter 6 has generated a huge class war between those of us who are fortunate enough to play with capital or technology and those of us who still hold a wrench. We see it in the campaign speeches. We see it on the television. And we saw it in the rejection of the rescue plan as a rich guy bailout.
I am very concerned that we have not seen the worst of this.
Nothing contained in this blog is to be construed as necessarily reflecting the views of the Pacific Research Institute or as an attempt to thwart or aid the passage of any legislation.