VoXEU.org, as the name suggested is a EU assisted portal set up by the Centre for Economic Policy Research (www.CEPR.org) in conjunction with a consortium of national sites. Vox aims to promote research-based policy analysis and commentary by leading scholars. But the website forces the authors to limit the commentary to maximum 1500 words a piece, making it accessible to a commoner.
There are perhaps a hundred explanations for the crisis in its aftermath (and before the crisis too, but who listens when the going is fine?) but a couple of ideas are worth highlighting.
Why did the Fed pursue a loose monetary policy for a long time?
It is said rightly that Fed kept a low interest rate regime for too long, contributing to asset bubbles. But why did it do that? Axel Leijonhufvud, Professor of Economics at UCLA, attributes the failure to inflation targeting, which is long heralded as the central bank's main job. In spite of the loose monetary policy, inflation stayed low for almost 5 years, because the developing economies kept their home currencies from appreciating and flooded the US markets with cheap imports.
Axel goes on to argue that while the imports kept the core inflation down, the Fed didnt recognise the asset price inflation it helped create(Alan Greenspan called it Asset froth instead of bubble, the typical George Orwell-coined-doublespeak).
In Age of Turbulence (the book was released when its author Alan Greenspan still had a reputation) Greenspan says how disturbed he was seeing the long rates (10 year GSec rates, the market's inflationary expectation) go down when the Fed started to tighten Monetary policy in June 2004 (He thought the fall was because of a global disinflation phenomenon, owing to rising productivity which put a lid on wages. He says and I quote, '..One recent evidence is the extraordinary number of labor contracts with 5-6 years maturities. We never had labor contracts of more than 3 years duration in the past 30-40 years').
To see how liquidity creates asset bubbles, we should recognise that financial market works differently from a Bread market, in that the demand and supply doesnt balance by price discovery, but the effects of high prices and high leverage are reinforcing. A typical investment bank had a capital of 1$ and borrowed $24 to buy assets worth $25. (the leverage of 24 might look eye-popping, but Lehman Brothers operated at a peak leverage of 32 in the First Quarter 2008). Assuming the assets earned 0.5% more than liabilities, the Return on equity was 12.5%. Since everyone chases the return, the spread of 0.5% narrows significantly, characterized by low risk premiums.
The only was to maintain RoE was to either hoard up more leverage, or chase riskier asset classes. Either actions have a reinforcing effect on prices. This is different from Bread market where the demand cools when prices rise.
Leverage works both ways- a drop in about 20% value in a portfolio where 20% of the assets are in MBS can wipe out capital (Case shiller home price index was down 23% from 2006 levels).
The solutions suggested aren't path breaking- more capital and more reserves to constrain leverage, but that's not my point. What do you do when you read two interpretations for the same phenomenon and both looks fine at that moment? Am I suffering something similar to Harry Truman's one-handed-economist syndrome?
Sunday, January 18, 2009
Wednesday, January 14, 2009
What is your most hated phrase?
The oxford list of most hated expressions must be dated. Here are the few economics jargons (although becoming layman expressions nowadays) I find annoying.
My top 10.
My top 10.
- Too big to fail (Best left unsaid)
- ____ gets worse before they get better (Fill in the blanks)
- Pump priming
- Stimulus (Even a half percent cut in excise duty of an esoteric product is being bandied as one)
- Animal spirits
- Greed and Fear
- Bailout
- Bottoming out
- Panic out there
- Monetary policy is losing traction
Tuesday, January 13, 2009
University of Chicago on Credit crisis II
As I said, the crisis can be broken into four parts, origins, Liquidity crisis, incentives and the policy responses. The first two were covered here.
Incentives
Creating liquidity in the otherwise illiquid subprime mortgage market, an otherwise noble goal, created perverse incentives for the lenders. Amit Seru, Professor at the University of Chicago did a research on the both the number of loans originated and the default rate around the FICO score of 620. FICO measures the credit worthiness of individual borrowers and a score of 620 and above is considered eligible for guarentee by Fannie/Freddie. An analysis of the origination of loans around 615-619 and about 620-624 showed a sharp spike in loan origination at around 620.
Since the lenders were eager to get the borrowers at that threshold, this jump could probably be explained. But what is more interesting is that the probability of default against the FICO scores. Normally it should have a negative slope- higher the score, lower the probability. What the results showed was that there was a jump in default at a score of 620, implying the lender did not do the due diligence because he knows that for scores above the threshold, the loans can be securitized and sold off. It could also mean that the borrowers know the threshold themselves and cheat their way to get just above 620 to qualify for a loan.
Assuming the second reason, while plausible but hard to detect, is not a major factor, it is safe to conclude that ceteris paribas (loan contract terms especially remaining the same), the incentives play an important part in due diligence.
Fiscal response to the crisis
Anil Kashyap, the delightfully articulate economist (Have you read this?) is an expert on Japan. He makes a convincing case for things to avoid in a response to the crisis which has an eerie similarity to the present one (Or for that matter, most crisis have the same cause- falling house prices- When would people learn that anything that rises can fall?) Japan is famous for its lost decade because the government failed to recapitalize the Banking system for a long time. The Government initiailly tried to deny the problem. They tried to hide the bad assets by creative accounting rules - Banks were allowed to chose which ones to carry at market values and which ones at book values!! In November 1997, when multiple large institutions failed (sounds like deja vu) the government got involved in half hearted recapitalization (The amount desired by the strongest bank was given to all banks as part of recapitalization)
These attempts shows valuable lessons for the current crisis. The strong banks are likely to ask the government to buzz off when offered capital, fearing Equity dilution of the existing shareholders. But it is advisable to recapitalize strong banks (or even encourage private funding). Its critical to stop dividend payments by the newly recapitalized banks, which are frankly money laundering of tax payer money. Better still, stop dividend payments done by all banks, strong and weak for 3 years, in order to nullify signalling effects associated with dividends.
The site is new but its well worth the read.
Incentives
Creating liquidity in the otherwise illiquid subprime mortgage market, an otherwise noble goal, created perverse incentives for the lenders. Amit Seru, Professor at the University of Chicago did a research on the both the number of loans originated and the default rate around the FICO score of 620. FICO measures the credit worthiness of individual borrowers and a score of 620 and above is considered eligible for guarentee by Fannie/Freddie. An analysis of the origination of loans around 615-619 and about 620-624 showed a sharp spike in loan origination at around 620.
Since the lenders were eager to get the borrowers at that threshold, this jump could probably be explained. But what is more interesting is that the probability of default against the FICO scores. Normally it should have a negative slope- higher the score, lower the probability. What the results showed was that there was a jump in default at a score of 620, implying the lender did not do the due diligence because he knows that for scores above the threshold, the loans can be securitized and sold off. It could also mean that the borrowers know the threshold themselves and cheat their way to get just above 620 to qualify for a loan.
Assuming the second reason, while plausible but hard to detect, is not a major factor, it is safe to conclude that ceteris paribas (loan contract terms especially remaining the same), the incentives play an important part in due diligence.
Fiscal response to the crisis
Anil Kashyap, the delightfully articulate economist (Have you read this?) is an expert on Japan. He makes a convincing case for things to avoid in a response to the crisis which has an eerie similarity to the present one (Or for that matter, most crisis have the same cause- falling house prices- When would people learn that anything that rises can fall?) Japan is famous for its lost decade because the government failed to recapitalize the Banking system for a long time. The Government initiailly tried to deny the problem. They tried to hide the bad assets by creative accounting rules - Banks were allowed to chose which ones to carry at market values and which ones at book values!! In November 1997, when multiple large institutions failed (sounds like deja vu) the government got involved in half hearted recapitalization (The amount desired by the strongest bank was given to all banks as part of recapitalization)
These attempts shows valuable lessons for the current crisis. The strong banks are likely to ask the government to buzz off when offered capital, fearing Equity dilution of the existing shareholders. But it is advisable to recapitalize strong banks (or even encourage private funding). Its critical to stop dividend payments by the newly recapitalized banks, which are frankly money laundering of tax payer money. Better still, stop dividend payments done by all banks, strong and weak for 3 years, in order to nullify signalling effects associated with dividends.
The site is new but its well worth the read.
Monday, January 12, 2009
Podcast, my new fad
Since I spend almost 3 hours a day commuting to work in the congested Mumbai roads, and speed read both my newspapers in half the time, I guess there must be something better than listening to the same songs in my ipod everyday.
I never realised all the news agencies/magazines/Business schools distribute surprisingly good analysis on variety of topics for free. Podcasts from Chicago Booth and Economist are very good. Bloomberg has some good pieces too.
I don't know how long this would last.
I never realised all the news agencies/magazines/Business schools distribute surprisingly good analysis on variety of topics for free. Podcasts from Chicago Booth and Economist are very good. Bloomberg has some good pieces too.
I don't know how long this would last.
Sunday, January 11, 2009
How do you forecast GSec rates?
Why do i care? Because my boss asked me for a simple model which predicts 10 year yields. Why do you care? Since you are reading my blog, may be you have some interest in the random thoughts running through my brain. So bear with me.
I did what everyone does. Google. The paper written by my Economics prof Rudra (say Rudro, he is very particular about the pronunciation) takes into all sorts of macroeconomic parameters, takes the data from a particular time period, generates the important factors that affect 10 year yields, creates a linear equation with a lag, makes an out of sample prediction for a period of 8 months (8 data points), calcualtes the RMPSE (root mean predictive squared error and compares that with RMPSE with a model generated with a trend line and says what the improvement is.
Cool. Next I asked our economist team here if they have done any econometric forecasting like that (I havent taken any econometric courses to replicate the method). Not only have they not done anything like that, but the guy says the forecasting of long term yields is rubbish, as the rates are entirely based on expectations and not based on history. In any case, the volatilty is unprecedented.
It is a universally acknoledged fact that NOW is always somewhat unprecedented. But as long as you take the right factors, you should still be closer to the truth. After all, expectations do not come from the heavens, but based on the data available. I would use my limited knowledge. All I want is a simple model.
I took exactly the same data (BSE100, REER, M3, WPI, Oil prices, IIP etc.) I dont have the tools to account for multicollinearity (two X factors are correlated) but I can still filter the factors based on t-statistics till all the factors in the regression equation have low p-values. Also, forecasting requires that I take all the inputs with a lag.
After 2 iterations, I filter down to just 3 factors - REER(t-1), IIP (t-1) and GSec (t-1). Based on the out of sample forecasts the actual and forecasted came close (forecast period Mar to Oct 08).
To be fair, the last month was unprecedented. The yield fell practically over 200 bps (from 7% in November end to low of 4.86% in December end) and then rose 120 bps last week. The fall of 200 bps was expected; The additional government borrowing of Rs. 5000 crores which pushed the yield up 120 bps was unexpected. Any model however correct (I asked the economist to check and if possible refine it) has to be supplemented with Qualitative data. Inspite of this, and the recent discredited models of credit rating agencies and Taleb's wide popularity, I think a model gives the relationship between variables elegantly in a way no amount of theory could.
I did what everyone does. Google. The paper written by my Economics prof Rudra (say Rudro, he is very particular about the pronunciation) takes into all sorts of macroeconomic parameters, takes the data from a particular time period, generates the important factors that affect 10 year yields, creates a linear equation with a lag, makes an out of sample prediction for a period of 8 months (8 data points), calcualtes the RMPSE (root mean predictive squared error and compares that with RMPSE with a model generated with a trend line and says what the improvement is.
Cool. Next I asked our economist team here if they have done any econometric forecasting like that (I havent taken any econometric courses to replicate the method). Not only have they not done anything like that, but the guy says the forecasting of long term yields is rubbish, as the rates are entirely based on expectations and not based on history. In any case, the volatilty is unprecedented.
It is a universally acknoledged fact that NOW is always somewhat unprecedented. But as long as you take the right factors, you should still be closer to the truth. After all, expectations do not come from the heavens, but based on the data available. I would use my limited knowledge. All I want is a simple model.
I took exactly the same data (BSE100, REER, M3, WPI, Oil prices, IIP etc.) I dont have the tools to account for multicollinearity (two X factors are correlated) but I can still filter the factors based on t-statistics till all the factors in the regression equation have low p-values. Also, forecasting requires that I take all the inputs with a lag.
After 2 iterations, I filter down to just 3 factors - REER(t-1), IIP (t-1) and GSec (t-1). Based on the out of sample forecasts the actual and forecasted came close (forecast period Mar to Oct 08).
To be fair, the last month was unprecedented. The yield fell practically over 200 bps (from 7% in November end to low of 4.86% in December end) and then rose 120 bps last week. The fall of 200 bps was expected; The additional government borrowing of Rs. 5000 crores which pushed the yield up 120 bps was unexpected. Any model however correct (I asked the economist to check and if possible refine it) has to be supplemented with Qualitative data. Inspite of this, and the recent discredited models of credit rating agencies and Taleb's wide popularity, I think a model gives the relationship between variables elegantly in a way no amount of theory could.
University of Chicago on Credit crisis
Initiative on Global markets have a surprisingly good 4 part lecture on Credit crisis each dwelling on one aspect of credit crisis - the origins, the liquidity crisis, incentives and the policy responses- lessons from Japan. While the conclusions reached are fairly conventional wisdom, the rigor of the research is something of note.
Mortgage crisis
Amir Sufi divides the areas into Prime and Subprime Zip codes (those areas where more than 60% of the loans originated by subprime borrowers are Subprime zip code areas while those with more than 60% prime are prime zip code areas. The growth in loans in subprime zip codes are about three times the growth in Prime areas between 2002 and 2005. There is a precedent in this even between 1999 and 2001, so he chose to see if there is any difference between then and now.
The factors why the loan growth exploded can be a) income growth in subprime is greater than prime areas; b) House price expectations were relatively higher in subprime and c) securitization was higher in subprime, with the associated incentives (as suggested by conventional wisdom)
If a) was indeed true, there seems to be a fair case for the growth in subprime loans. In fact between 1999 and 2001, that was indeed the case. Between 2002 and 2005, the growth in income was about 4% for subprime while for prime borrowers was 8%. Even then, if there is a possibility that the subprime borrowers crossed a threshold to qualify for a homeloan so the home loans exploded (its a similar case with demand for cars in India- a once an income threshold is crossed, the market explodes) So Amir takes only the zip codes where there is a negative growth in nominal income and sees the relation between loan origination growth between subprime and prime growth. Even there, the loan growth is starkly higher.
So the second explantion for the loan growth, higher house price appreciation for subprime borrowers does have some evidence. But it may be the case that the price appreciation happened precisely because of loan origination growth. It is difficult to disentangle the effects between the two variables when causality is the case. Which makes it all the more curious that rating agencies put the house prices on the RHS of the equation for giving the ratings for securitized debt.
There is also strong evidence for the third relationship but it is taken up in the third part. The moot point is micro trends are important for predicting a crisis- how did loan origination grow faster in a segment which has seen slower or even negative growth in income compared to prime borrowers, as was seen as early as 2004? The other conclusion is its important to not treat house prices as exogenous.
Liquidity crisis
Doug Diamond says Banks, investment banks and hedge funds are by nature, highly levered institutions. It is impossible for investors to check the quality of assets of a bank as compared to say, a car company. Which is why both Equity and long term debt is in short supply for a bank and they rely on deposits or wholesale funding.
The more difficult it is to judge the quality of assets, the more levered the institution. That is why Bear Sterns or Lehman have leverage of about 30-35 while even Goldman Sachs had to be satisfied with a modest 20 times leverage. And the funding for the investment banks is almost entirely overnight, because they cannot monitor what the banks do with their money even less than a commercial bank. The investment banks can rapidly alter the risk profiles on a daily basis since they have a trading portfolio as compared to a loan portfolio of a commerical bank whose risk profile is sticky. In this scenario, the only bargaining chip the investors have is that they can stop rolling over short term money. It keeps the investment banks in check, but it also increases the risk of run, which is costly for both the borrower and the lender. If it is costly, why does the investor do it? In times of uncertainity, the investor knows if he doesnt pull out, someone else will and they would get the 100 cents on dollar while he potentially loses some because he was patient.
This is what happened with Northern Rock, a fundamentally solvent bank with no exposure to US subprime but got its funding mostly from ABCP market. With that market crashing because investors stopping to rollover funds for anyone with any exposure to Mortgage assets (Northern Rock assets were mostly in UK prime segment). The fear of solvency was enough to start a run on the bank which became a self fulfilling prophecy.
When the bank fears a run because the capital has become low, it can either raise capital or dump assets to bring down the leverage. Banks usually opt for the second because the first may take time or prove difficult. But there was no market for these assets because of two reasons. First, if the investors think if they dont buy assets today worth $5 today they can get it even cheaper tomorrow (perhaps $2) they would stop buying it. Two, if other instituions think if they buy these assets cheap, they have to mark their similar investments down which results in their capital levels getting low, they wouldn't buy it. This is what TARP 1 wanted to correct; Paulson thought if the government becomes the buyer of dodgy investments of the last resort, the lower bound can be made $5 or even the hold to maturity value of the asset, so that the actual payout by the government may not happen at all, and the instituions may appear solvent. But the more direct approach is to direct equity into the troubled banks so that they need not firesale the assets at all. TARP 1 went into rough weather.
The lecture discusses the effect of short term debt remains the same for every crisis, right from Bank runs of 1930s to current one. The solution proposed then was to insure all short term debt for 90 days, conduct audits for all banks, differentiate good banks from bad, inject equity into good banks, merge the average ones, and let the bank banks go through resturcturing through the FDIC route. Now that the liquidity crisis have largely eased as the Fed is lending to everyone, we may not know how that would've worked.
The other two aspects we can see later.
Mortgage crisis
Amir Sufi divides the areas into Prime and Subprime Zip codes (those areas where more than 60% of the loans originated by subprime borrowers are Subprime zip code areas while those with more than 60% prime are prime zip code areas. The growth in loans in subprime zip codes are about three times the growth in Prime areas between 2002 and 2005. There is a precedent in this even between 1999 and 2001, so he chose to see if there is any difference between then and now.
The factors why the loan growth exploded can be a) income growth in subprime is greater than prime areas; b) House price expectations were relatively higher in subprime and c) securitization was higher in subprime, with the associated incentives (as suggested by conventional wisdom)
If a) was indeed true, there seems to be a fair case for the growth in subprime loans. In fact between 1999 and 2001, that was indeed the case. Between 2002 and 2005, the growth in income was about 4% for subprime while for prime borrowers was 8%. Even then, if there is a possibility that the subprime borrowers crossed a threshold to qualify for a homeloan so the home loans exploded (its a similar case with demand for cars in India- a once an income threshold is crossed, the market explodes) So Amir takes only the zip codes where there is a negative growth in nominal income and sees the relation between loan origination growth between subprime and prime growth. Even there, the loan growth is starkly higher.
So the second explantion for the loan growth, higher house price appreciation for subprime borrowers does have some evidence. But it may be the case that the price appreciation happened precisely because of loan origination growth. It is difficult to disentangle the effects between the two variables when causality is the case. Which makes it all the more curious that rating agencies put the house prices on the RHS of the equation for giving the ratings for securitized debt.
There is also strong evidence for the third relationship but it is taken up in the third part. The moot point is micro trends are important for predicting a crisis- how did loan origination grow faster in a segment which has seen slower or even negative growth in income compared to prime borrowers, as was seen as early as 2004? The other conclusion is its important to not treat house prices as exogenous.
Liquidity crisis
Doug Diamond says Banks, investment banks and hedge funds are by nature, highly levered institutions. It is impossible for investors to check the quality of assets of a bank as compared to say, a car company. Which is why both Equity and long term debt is in short supply for a bank and they rely on deposits or wholesale funding.
The more difficult it is to judge the quality of assets, the more levered the institution. That is why Bear Sterns or Lehman have leverage of about 30-35 while even Goldman Sachs had to be satisfied with a modest 20 times leverage. And the funding for the investment banks is almost entirely overnight, because they cannot monitor what the banks do with their money even less than a commercial bank. The investment banks can rapidly alter the risk profiles on a daily basis since they have a trading portfolio as compared to a loan portfolio of a commerical bank whose risk profile is sticky. In this scenario, the only bargaining chip the investors have is that they can stop rolling over short term money. It keeps the investment banks in check, but it also increases the risk of run, which is costly for both the borrower and the lender. If it is costly, why does the investor do it? In times of uncertainity, the investor knows if he doesnt pull out, someone else will and they would get the 100 cents on dollar while he potentially loses some because he was patient.
This is what happened with Northern Rock, a fundamentally solvent bank with no exposure to US subprime but got its funding mostly from ABCP market. With that market crashing because investors stopping to rollover funds for anyone with any exposure to Mortgage assets (Northern Rock assets were mostly in UK prime segment). The fear of solvency was enough to start a run on the bank which became a self fulfilling prophecy.
When the bank fears a run because the capital has become low, it can either raise capital or dump assets to bring down the leverage. Banks usually opt for the second because the first may take time or prove difficult. But there was no market for these assets because of two reasons. First, if the investors think if they dont buy assets today worth $5 today they can get it even cheaper tomorrow (perhaps $2) they would stop buying it. Two, if other instituions think if they buy these assets cheap, they have to mark their similar investments down which results in their capital levels getting low, they wouldn't buy it. This is what TARP 1 wanted to correct; Paulson thought if the government becomes the buyer of dodgy investments of the last resort, the lower bound can be made $5 or even the hold to maturity value of the asset, so that the actual payout by the government may not happen at all, and the instituions may appear solvent. But the more direct approach is to direct equity into the troubled banks so that they need not firesale the assets at all. TARP 1 went into rough weather.
The lecture discusses the effect of short term debt remains the same for every crisis, right from Bank runs of 1930s to current one. The solution proposed then was to insure all short term debt for 90 days, conduct audits for all banks, differentiate good banks from bad, inject equity into good banks, merge the average ones, and let the bank banks go through resturcturing through the FDIC route. Now that the liquidity crisis have largely eased as the Fed is lending to everyone, we may not know how that would've worked.
The other two aspects we can see later.
Friday, January 09, 2009
Firefox is better than chrome
I had stopped using firefox when google released the beta version of chrome. But now I realize its still some way to go in terms of speed and performance.Some pages take eons to load and sometimes it just gets hung. That every window is a seperate process is only quantum of solace. What say?
Wednesday, January 07, 2009
What does your blog say about you?
Typealyzer.com apparently answers the question with a warning- your writing style may have nothing to do with your self perceived personality.
I thought I would do better than a mechanic. Heck, even Paul Krugman and my friend are mechanics.
I thought I would do better than a mechanic. Heck, even Paul Krugman and my friend are mechanics.
Satyameva Jayate (Truth shall triumph)
If you see one cockroach scurrying in the light, there are probably a dozen in the shade. When Satyam (Sanskrit word means truth) board accepted a proposal to buy the family firm (Maytas, no prize for guessing the similarity) of promoters for a whopping $1.6 billion, there was so much hue and cry with people wondering why the board members (including the independent directors) bowed to one man who had barely 5% stake in the company. As it turns out, they were indeed acting in the best interests of the company. An overpaid asset is worth more than a fictitious one. So much for shareholder activism.
It might seem like a double whammy that just when the apocalypse predictions about the world are coming true, there is a big wave of frauds hitting the markets. But as Warren Buffett once said, "You only find out who is swimming naked when the tide goes out". If luck can run out for Bernie Madoff after running a con operation for decades, what hope does Raju (Satyam chairman) have? In fact, there are reports that SEC was indeed warned about scandal as early as November 2005, full two years before the revelation (?!) Who knows if Krishna Palepu, the esteemed Harvard professor, independent director at Satyam and a leading authority on Corporate governance knew this all along? Or for that matter, how did PwC audit a non-existent cash and Bank balances?
Eventually, Satyameva Jayate. But be wary when you switch on the light just yet.
It might seem like a double whammy that just when the apocalypse predictions about the world are coming true, there is a big wave of frauds hitting the markets. But as Warren Buffett once said, "You only find out who is swimming naked when the tide goes out". If luck can run out for Bernie Madoff after running a con operation for decades, what hope does Raju (Satyam chairman) have? In fact, there are reports that SEC was indeed warned about scandal as early as November 2005, full two years before the revelation (?!) Who knows if Krishna Palepu, the esteemed Harvard professor, independent director at Satyam and a leading authority on Corporate governance knew this all along? Or for that matter, how did PwC audit a non-existent cash and Bank balances?
Eventually, Satyameva Jayate. But be wary when you switch on the light just yet.
Subscribe to:
Posts (Atom)
