Tuesday, January 23, 2007

Trends in Central Banking

What is the Federal Reserve?
Of the US population
23% think it was an Indian reserve,
26% a wildlife preserve,
51% a brand of whiskey
Remarks of Chairman Miller in Peter Lynch’s One up on Wall Street

Dear readers of The Jag! This edition of the ‘newsletter with teeth’ features an article on central banking that was written around the end of March 1999. We are reproducing it in here as a background article for further analysis that will be featured in forthcoming editions. Happy reading and stay tuned!
__________________________________________________

Central Banks could probably have done without the Y2K problem. The 90s would already have been an unusually busy decade for many of them. Here is how.

Inflation targets, uncertainty and the Monetary Conditions Index
Early on in the decade, mounting deficits and ballooning debt levels in G-7 countries have forced central banks and governments there to opt for low inflation targets. Roughly, between 1% and 5% as captured by the CPI[1]. It’s not that it suddenly dawned upon them that interest payments were eating up 30 cents out of every tax dollar. No. It was more because of a wake up (and smell the coffee!) call from our friendly global capital markets telling them to either put their fiscal houses in order and adopt consistent monetary policies or face a massive sell off of their securities. Some like Canada have listened and have found many virtues to low inflation (which its Central Bankers are likely to have happily shared once a month with their colleagues in Basel, Switzerland) along what was to be a very bumpy road. The good news is that they will turn out a budget surplus this year from a deficit of CAD 42 billions in 1993. Finance Minister Paul Martin, a millionaire from Montreal and strong contender to succeed Jean Chretien as PM, will use part of the surplus to start paying off the national debt.

Why adopt low inflation targets?
First because it shields the purchasing power of pension benefits of a rising share of old folks. It also bolsters confidence in the currency. Next it allows economic agents to get their priorities right by focusing on productivity-enhancing vs. speculation activities like those experienced in Germany in the 30s. Collective bargaining is carried out in an environment of lower inflation uncertainty and is likely to lead to a lengthening of contract maturities. Lower inflation will also ensure lower borrowing costs via lower inflation premium and lower inflation variability (in a GARCH sense). Once inflation is tamed, rates should come down, increasing investments and the general standard of living.

Central banks have not only to reduce year-on-year changes in the CPI index but they also have to drive down expectations of inflation. And they have to do that throughout the whole business cycle. Low inflation credibility will not be sacrificed at the expense of averting an economic slowdown. Besides as Central Bankers have found out, it takes a while for credibility to show up in a markov-switching model (the new darling in central bank research).

How uncertainty can being reduced
By establishing price stability as a goal. The name of the game has been first a systematic reduction in the rate of inflation and the subsequent adoption of specific inflation targets. This is the approach adopted by the Reserve Bank of New Zealand. Once this is achieved, the Central Bank has to stay the course and possibly focus the public’s attention on its use of intermediate keyposts such as the Monetary Conditions Index (MCI) in gauging the stance of monetary policy. Central banks around the world have been making a greater effort to communicate policy actions and intentions through press releases, monetary policy reports, press conferences and of course via their websites.

The MCI
Pioneered by the Bank of Canada, the MCI is a tracking device that is correlated with the ultimate target such as aggregate demand while being sensitive to policy instruments. It is a combination of interest rates and an index of exchange rates that reflects the relative importance of these two channels of monetary policy in affecting nominal spending. A typical vector autoregression result is that interest rates are between 2 (New Zealand) to 3 (Canada) times more important than exchange rates. Which is good news for Central Banks as they have little influence over the exchange rate when their currencies float. The MCI has also found a place in the toolbox of the quantitative asset manager namely in the design of asset allocation models.

Problems with the CPI and the monetary aggregates
The recipe appears simple. Reduce the monetary ease and this will bring the CPI within the targeted band. The snag is that the CPI contains a measurement bias. Reasons advanced by Federal Reserve officials in their testimonies are that current inflation rates imply that there has not been any increase in productivity, the relevant baskets are slow to acknowledge technological advances that modify consumer behaviour and yield productivity gains. The Fed estimates that inflation is overstated by 1% to 1.5%. This makes the case for a positive rate of inflation. To complicate matters growth appears to be overstated by 0.5%. Solutions include the adoption of chain weighted indices, a subjective removal of the bias and developing better methodologies for measuring increases in the general level of prices.

Another show-stopper is that the link between M1 and nominal spending has weakened enough to prompt Feds to reduce its importance among the battery of indicators or abandon it altogether (US, Canada while the Bundesbank watched M3 in the pre-Euro setup). The idea is to identify broader monetary aggregates that provide good informational content of inflation 4-8 quarters hence as monetary policy operates with a lag.

Banking supervision and the management of reserves portfolios
Barings, Credit Lyonnais, Orange County, Daiwa and more recently LTCM. These are financial disasters that had the potential to torment bank regulators in their crusade of ensuring a healthy financial system. These events have since led to the establishment of capital adequacy standards with respect to market risks over and above those that have traditionally been required as a buffer against unpleasant credit risk surprises. But financial institutions have recently been given the option of utilizing their own risk management systems in determining the amount of capital to set aside for these risks. This is essentially because the risk management culture at some of the bigger and more innovative banks have been found to be more sophisticated than anything regulators could have asked for or imagined. One example is JP Morgan’ s 4:15 report that spits out their forecasted global market risk on a daily basis.

And Central Banks are integrating some of these ideas in the active management of the reserves portfolios. Gone are the days of passive management of short-dated instruments. Duration gap models are being replaced by value-at-risk[2] metrics coupled with stress simulations of aggregate market risks. They have found that drawing on the seminal work of Markowitz, Sharpe, Merton and the likes are not inconsistent with their basic objectives. Furthermore optionality at the security and portfolio level matters. And they too are worried that asset and default correlations tend to increase when they are needed most. In sum they want to make sure that the return earned (contribution to the national till) is enough for the level of risk assumed. You wouldn’t want it to be otherwise.

On independence
This is a concept espoused by the Bundesbank in 1924 when the relevant legislation was enacted. Since then Federal elections have been lost and won over the single issue of price stability. The relationship between independence and the inflation record of the Central Bank has been investigated too. The GMT index, a weighted sum of 15 legal provisions that includes the length of the governor’s contract and whether he is a government appointee, has been found to have good predictive value. We can only hope that country-rating agencies will factor this into their analysis of our island once they spot Mauritius on a world map that is.

Implications for the investor
Central banks can have an enormous influence on financial markets. Let us consider fixed-income markets. The Treasury yield curve is the benchmark for corporate debentures (bonds) everywhere in the world. Every corporate issue is priced off this curve as it is considered risk-free. In Mauritius, the curve extends out to a maturity of about two years thanks to the seven-hundred-and-something-day T-bill. Corporate debentures of similar effective maturity (which is the case of all but three debentures) should offer a higher yield than their Treasuries counterpart because of default risk, redemption risk, redemption behaviour, horizon risk and reduced liquidity that characterise these issues. The impact of tax-free interest income should likewise be taken into account. All these factors along with the availability of T-bills to the individual investor explain why corporate debentures appear to trade a bit more consistently with respect to government paper. A closer analysis is however required to find out where they are heading to. Empirical research in the US reveal that the influence of the Fed via the fed funds market on the Treasury yield curve is concentrated mostly at the shorter end, accounting for more than three quarters of the variability of changes in T-bill rates whereas it explains about one only third of the movements in 10-year bonds.

[1] Or core inflation. That is CPI without its volatile components such as food and energy components and indirect taxes. But inflation targets can be more sophisticated. Rules like a 3% annual increase in the monetary base with adjustments for the change in base velocity over the past four years plus for deviation of nominal GNP from target paths have been suggested.
[2] The US Securities and Exchange Commission now requires that all companies with a market capitalization of $2 billion and more (about 25% more than the market cap of the SEM at the time of writing) to disclose any material risk exposures.
Comments: density@intnet.mu.
No. 7 January 2007
© Sanjay Jagatsingh, 2007