Tuesday, December 29, 2009

Process versus Outcome

We are coming to the end of the first decade of the millennium. If I have to think of one lesson that I have learnt over the last decade, then it has to be the one on 'process versus outcome'.

To illustrate this concept, let me cite an example from Michael Lewis' book Moneyball:

Paul DePodesta, a former baseball executive tells about playing blackjack in Las Vegas when a guy to the right, sitting on a seventeen, asks for a hit. Everyone at the table stops, and even the dealer asks if he is sure. The player nods yes, and the dealer, of course, produces a four. What did the dealer say? "Nice hit." Yeah, great hit.

Similar to this anecdote, in our daily lives, we measure performance by results. After all, results are what ultimately matter - they add to bottom line. Evaluating the underlying process is subjective and hard. So, we simplify our lives, by making the critical mistake of assuming that good outcomes are the result of a good process and bad outcomes are the result of a bad process. Besides, the only ones voicing against doing so are mostly people who have 'failed' in their endeavours.

But alas, reality is far more vicious than a game at Vegas. We all wish life was as simple as a game of blackjack where each card about to open up is one of the few known options. But often we are faced with a myriad of possibilities. Our folly is in picking a process that has the possibility of delivering a rare but fatal outcome. On top of that, a few repetitions of the process starts giving us a false sense of security - we start feeling that the sorts of things that happen to others will not necessarily happen to us. And then if nothing bad happens then we kick our self for being too worried.

Nassim Taleb, the author of Fooled by Randomness, offers a thinking tool called 'alternate histories' that I find very useful:

The concept of 'alternate histories' is similar to the many-world interpretation in quantum mechanics, which considers that the universe branches out treelike at every juncture; what we are living now is only one of these many worlds. Taken at a more extreme level, whenever numerous various possibilities exist, the world splits into many worlds, one world for each different possibility - causing the proliferation of parallel universes.


If we apply the 'alternate histories' way of thinking then the guy at the blackjack table got busted in one of the alternate worlds indicating that he was following a poor process. Even though this way of judging matters is abstract and counter intuitive, it is a good way to evaluate our process. After all, in the long haul, a good process gives us the most reliable way of raising our chances of a good outcome.


  • Good process, Good outcome - Deserved success
  • Good process, Bad outcome - Bad luck
  • Bad process, Good outcome - Dumb luck
  • Bad process, Bad outcome - Poetic Justice

Monday, December 28, 2009

Playing by the (Basel) Rules


US private debt grew from 20 trillion dollars in 2000 to over 50 trillion dollars in 2007, but FDIC regulated banks (referred to as banks for the rest of the discussion) continued to stay "well capitalized" during this period. How could total debt grow by 80% without leading to the deterioration of bank's capital?

To answer this question, we need to look at how banks are regulated. At the end of 1980s, the G-10 countries decided to coordinate their banking regulation through the Basel Committee on Banking Supervision of the Bank for International Settlements (B.I.S.) by setting capital adequacy requirements. According to the Basel rules, all banks and depository institutions in the countries that adhered to them must maintain a certain minimum fixed amount of capital in relation to its assets.

The capital adequacy requirement set by B.I.S is to protect the banks from unexpected losses, since the banks are protected by expected losses by accounting for them on their books. The term "capital" might give you the impression that it is cash held by the banks in their vaults. But such a usage is misleading. Generally speaking, capital is the portion of bank's assets that don't have to returned to creditors (depositors are also creditors). It is only because of the fact that this portion of its financing does not have to be repaid that the bank has the capacity to withstand unexpected losses. It is the capital that absorbs these unexpected losses. The Basel rules classify capital into 2 tiers - Tier 1 "core" capital and Tier 2 capital. Tier 1 capital largely consists of funds raised through selling common stock, disclosed reserves, and retained earnings. Tier 2 is defined as undisclosed reserves, revaluation reserves, loan-loss reserves, convertible bonds, cumulative preferred shares, and subordinated debt.

The Basel rules require that the banks hold certain minimum ratios of clearly defined capital (calculated at book values) against assets that are adjusted by clearly defined weights.

Capital ratio = Capital / Risk-weighted Assets

Under this framework, banks to be considered "capitalized" are required to hold no less than 8 percent capital against total risk-weighted assets.

The system of assigning weights to assets is fairly standardized requiring minimal supplementation by various national banking regulators. In the United States some of the risk weighting rules are as follows (for details look at BIS 1988):


  • 0% weight to cash, gold, and bonds issued Organization for Economic Co-operation and Development (OECD) governments

  • 20% risk weight for AAA and AA rated asset-backed securities and claims on OECD banks, local public-sector entities, and agencies of OECD governments, such as the government sponsored enterprises Fannie Mae and Freddie Mac.

  • 50% risk weight to mortgage loans

  • 100% risk weight to all claims on the private sector and non-OECD governments, to investments in real estate, equities, corporate bonds, and all other assets rated lower than AA


In addition to the Basel rules, the FDIC in the United States require that banks aspiring to be deemed "well capitalized" - and thus enjoy valuable privileges like securities underwriting - must hold their capital in a configuration that meets additional three additonal ratios: capital to risk-weighted assets of 10 percent, tier-1 capital to risk-weighted assets of 6%, and tier-1 capital to total assets of 5%. The chart above plots these three ratios for American FDIC banks leading upto the financial crisis. It should strike you that, as per the Basel rules, banks were a few percentage points higher than those mandated by FDIC for being "well capitalized". This might seem hard to square with the expansion of indebtedness that took places during the credit boom years.

The regulatory ratios can be achieved by either increasing the numerator or by decreasing the denominator - by building up capital, or by cutting back on lending or reducing the riskiness of the balance sheet. But, with the credit boom underway, neither might seem feasible. But in fact, both methods were used extensively, which led to the impression of a financial system that was safer than it really was.

To illustrate the kind of games that the banks played, lets use an example offered by Robert Merton:

If a bank were managing and holding mortgages on houses, it would have to maintain a capital requirement of 4%. If, instead, it were to continue to operate in the mortgage market in terms of origination and servicing, but sells the mortgages and uses the proceeds to buy U.S. government bonds, then under the BIS rules, the US government bonds produce no capital requirements and the bank would thus have no capital maintainance. However, the bank could continue to receive the economic equivalent of holding mortgages by entering into an amortizing swap in which the bank receives the total return on mortgages, including the amortizing features and prepayments, and pays the return on US Treasury bonds to the swap counterparty. The net of that series of transactions is that the bank receives the return on mortgages as if it had directly invested in them. However, the BIS capital rules, instead of being 4 percent, apprears to produce a capital requirement using the swap route of only 0.5 percent.

Central to this example is the active management of a bank's balance sheet by selling and swapping assets through securitization. Playing this game is also called as regulatory arbitrage - restructuring a bank's portfolio so that it has the same or even greater risk as before, but a lower capital requirement.

It is costly to maintain capital - it lowers the profitability of the bank and constraints its growth. Thus, it is in the bank's interest to not hold mortgages on its books, but to transfer the mortgages to a securitizer such as Fannie Mae or Bear Stearns or "hide" it in its shadow bank conduit, freeing up the its capital. The freed up capital can either be used to pay down its debt (not what happened during the credit boom) or to expand its balance sheet by making more loans (what happened during the credit boom).

This is how banks continued to appear well capitalized and not reflect the economic reality of an indebted economy

Saturday, December 26, 2009

Investor's Myth: 'Higher the risk, higher the return'

I had a party at my place during the holidays and the men starting talking about investments they have been making. One of the guests started describing his success with speculative investments in the currency market and ended by saying: "The higher the risk you take, the higher the return". This is one of the most often repeated maxim (and I am about to argue it to be a false) in finance, but having read "How to win friends and influence people", I played the friendly host by avoiding an argument and politely changing the topic to the appetizing food at the party.

But I can make my case on my blog (that has very few readers and hopefully not the ones that came to the party). The notion of risk is one of the most misunderstood concepts in finance. So, what is risk? If you went to University of Chicago (I love my Chicago MBA friends - no offense to you guys), risk is explained using the capital asset pricing model (CAPM). In the case of the equity market, this risk is quantified using a statistical concept called the beta. As an example, lets look at the stock market. By defintion, the market is considered to have a beta of 1.0, and individual stocks in the market are ranked according to how much they deviate from the market. So, higher-beta stocks are the ones that are more volatile, and are considered 'riskier', according to CAPM. According to this theory, 'riskier' (high-beta) investments should have high long-term returns. And hence the statement: 'Higher the risk, higher the return'.

However, for this theory to be true, investors have to demand higher returns from 'riskier' (high-beta) stocks. There have to be people that demand such a relationship. In practice, there aren't many people who demand such a pricing, because most participants understand that risk is not the same as volatility. There are various types of risk: business risk (possibility of detoriation of operations, profit margins etc), financial risk (catastrophe due to high leverage), valuation risk (stock priced much higher than underlying intrinsic value of the business leaving little margin of safety), liquidity risk (inability to sell the stock in the market without affecting the price), and volatility risk (market price fluctuation). Off all these risks, volatility risk is of much smaller importance to most participants in the equity market. Besides, everybody knows that past results bear little resemblance to future results. So, why should historical volatility matter much to the future of the enterprise.

Since volatility is poor measurement of risk, lets replace the word risk with business risk and re-examine the maxim: 'Higher the business risk, higher the return'. Obviously, riskier (defined as business risk) investments cannot be counted on to deliver higher returns. Because if that were the case, then there is nothing risky about the investment. The correct formulation is that in order to attract capital, riskier investments have to offer the prospect of higher return. But there is absolutely nothing to say that these prospective returns will materialize.

Often, this simple logic is forgotten. Here is what really happens: Riskier investments are priced to deliver higher returns (if the investments materialize). Such a pricing is required to start attracting capital. In some cases, these investments start paying off handsome returns. The maxim 'Higher the risk, higher the return' starts getting repeated. And, hence the investment attracts more capital and bids up the price of such investments. At some point, the pricing of these investments are bid-up to such an extent that the investor is not compensated adequately to take the risk. Warren Buffet often says: 'What the wise do in the beginning, the fools do in the end'.



Here is a specific example from the recent credit boom and bust. At the begining of 2003, credit spread for high-yield bonds (another name for junk bonds) was at historical highs of over 1000 basis points. Investors were getting paid a premium of 10% over the default risk-free US treasury bonds to take the (credit, liquidity, market) risk of investing in junk bonds. In 2005, the investors were paid a mere premium of 200 basis points.

Here is another example. Cisco's stock was priced at 77$ in its peak in 2000 and had earnings of 0.36$. So, the investor in Cisco's stock in 2000 was paying 213x for the 0.36$ of earnings it had at that time. That is the equivalent of a mere 0.46% yield. Risk-free US treasuries were yielding over 6% at this time. This is yet another example of inadequate compensation of taking the risk of investing in the Cisco stock. Investor's were so seduced with the story of the internet boom that they overbid the price of the Cisco stock to an extent were they were paid nothing to take on the risk. We all know what happened in the end. Those who invested in the stock in 2000 are still waiting to make their return and this is from a successful enterprise that makes really phenomenal products



Investments are not risky inherently but only in relation to its market prices. So, the next time someone recommends a 'risky' investment as one with higher return, the first question you want to ask yourself is whether the investment is priced appropriately to compensate you for the risk you are about to take.

Thursday, December 24, 2009

Was the U.S. monetary policy too loose during the housing boom?

John Taylor is an economics professor at Stanford University. In a 1993 paper, he introduced the Taylor rule, which provides guidance to central banks on how to determine short-term nominal interest rate (called the federal fund's rate in the US). It relates the interest rates to the amount of slack in the economy and the inflation rate.

He presented at the Fed's annual conference at Jackson Hole in 2007 evidence that suggested that the Fed's loose monetary policy in the 2000-2006 period was too loose. He then uses this data to argue that it was one of the biggest triggers of the housing boom in the US.

The Economist published an article, Fast and Loose, in Oct 2007 that illustrates the monetary excesses. Below is a chart from the article.



The dot-com bubble had burst and the economy was in recession. By 2003, Mr Greenspan, the chairman of the Fed then, had lowered the federal fund's rate to 1%, the lowest since 1958. He kept the interest rate at 1% for an entire year. He justified this decision by saying that he feared that America was on its way to deflation. As per Taylor's rule, the interest rate should have been above 3%. So, even though the Taylor rule is only a guide, Mr Greenspan had missed the mark by about a mile. Furthermore, Mr Greenspan said that interest rate would be low for "a considerable period" and that the Fed would rise it slowly at a "measured pace".

Gradually the Fed started raising the interest rates in quarter point steps. By June 2006, the interest rate was at 5.25%, as recommended by the Taylor's rule. So, looking back at the chart, US had over 4 years of loose monetary policy - enough time for an asset bubble to grow - all in the fear of deflation.

The housing bubble has burst. And coincidentally, Mr Bernanke, the current chairman of the Fed, just repeated Mr Greenspan's 2003 words - low interest rates for a "considerable period" and Fed will raise the interest rate slowly at a "measured pace". Would it be a coincidence that the current loose monetary policy is giving birth to yet another asset bubble?

Tuesday, December 15, 2009

LIBOR-OIS spread

The Libor-OIS spread is an important barometer of stress in the banking system. The term London interbank offer rate (LIBOR) is the rate at which banks indicate they are willing to loan to other banks for a specified term of the loan. The term overnight indexed swap (OIS) rate is the rate on a derivative contract on the overnight rate (In the U.S. the overnight rate is the effective federal funds rate.) In such a contract, two parties agree that one will pay the other party a rate of interest that is the difference between the term OIS rate and the geometric average of the federal fund's rate over the term of the contract. Thus, the term OIS rate is the market's expectation of the federal fund's rate over the term of the contract. There is very little risk in the OIS market because there is no exchange of principal; funds are only exchanged at the end of the contract, when one party pays net interest obligation to the other party. The term Libor-OIS spread is be a measure of the health of the banks because it measures what banks believe is the risk of default associated with lending to other banks. Changes in Libor-OIS spread reflect changes in risk premiums and liquidity premiums.