This paper describes why we look for a less efficient calibration of the business cycle—maintaining inflation at target should become increasingly challenging for central bankers as shocks ripple through domestic economies unrestrained. We conclude that the road ahead will be bumpier than the road travelled, and investors may wish to acquire a high-quality set of shock absorbers. The recent spike in inflation volatility is a cyclical phenomenon that will recede (Exhibit 1). Nevertheless, we believe it is the beginning of a more subtle uptrend over the coming business cycle. We are not ready to dismiss the Great Moderation as a temporary illusion but an increased allocation to inflation-linked debt can improve the efficiency of most investors’ portfolios by absorbing many of the shocks that we expect to continue.

The 21st Century Financial System (or the Inefficiency of Policy Instruments)
A tenet of 20th Century monetary theory is that actual policy rates matter little; of paramount importance is the management of expectations of future policy rates. Households and businesses do not borrow from central banks in the overnight interbank market. Rather, borrowing decisions are based on five-, 10-, and 30-year interest rates, similar horizons of the activities being financed. Former Federal Reserve (Fed) Vice-Chairman Alan Blinder captured it best when he stated, “Central banks generally control only the overnight interbank rate, an interest rate that is relevant to virtually no economically interesting transactions.”¹

This tenet is due for revision. The rise of market-based shadow banking and the displacement of traditional deposit-taking banks have invalidated Blinder’s statement (Exhibit 2). The 21st Century financial system has become more collateral based. That is to say, more and more lending takes place against the pledging of collateral, in the form of securities or physical assets. In contrast, the 20th Century financial system was dependent not on collateral but instead on the onerous task of screening and monitoring in an attempt to uncover underlying trustworthiness. This change may have complicated a monetary transmission channel that was thought to have been fairly well understood and measured. Two alterations have made central bankers’ jobs of keeping inflation at target more challenging.

Alteration 1: The Long and Short of It
Short rates (uncollateralized LIBOR-based borrowing, or, more often, collateralized repurchase agreements) fund market-based shadow banking securitization activities, which have become a critical driver of economic activity. Short rates closely follow policy rates and now play a direct role in the monetary transmission mechanism, independent of their influence on long rates. Due to the shadow banking systems’ rise to prominence, central bankers must now simultaneously optimize the level of short- and long-dated rates in an effort to manage inflation pressures, all the while guessing at the relative influence of each. The slope of the yield curve is obviously taken into consideration. But central banks have never directly targeted both short- and long-dated rates independently, that is until recent quantitative easing programs were incepted.

Counterintuitively, collateralized funding can be even more volatile than uncollateralized LIBOR-based funding. The interest rate paid on funding via repurchase agreements, which are collateralized by informationally-insensitive debt (Treasuries, agencies and even highly diversified securitized debt), is partly determined by the size of the haircut. In addition to the cost of funds, the size of the haircut rises along with the volatility of the underlying collateral. As the collateral behind a repurchase agreement becomes less stable—or more informationally-sensitive—the haircut increases to offset these risks. So, for example, as collateral switches from short-dated T-bills to 30-year Treasury bonds, the size of the haircut increases to offset the higher duration (price sensitivity to interest rates) of the 30-year bond. And as collateral switches from Treasuries to assets with default risk, such as corporate bonds or even equities, the size of the haircut rises even more. Professor Gary Gorton of Yale has made an important contribution, estimating that the average haircut rose from near 0% in 2007 to almost 50% in 2009. In other words, as the Fed funds rate rose from 1% to 5.25% and LIBOR rose from 1% to 6%, haircuts on repo transactions rose from 0% to 50% (Exhibit 3). This was devastating to monetary conditions given a financial system that had come to rely so heavily on collateralized lending. Inferring the degree of monetary tightening by observing the level of longer-dated rates would have been dangerously misleading.

With this in mind, Alan Greenspan may have mistakenly focused too much attention on the “conundrum,” when long-dated yields declined in spite of rising policy rates. He was concerned that globalization had weakened the Fed’s influence, a sentiment echoed by The Economist, which wrote, “The yield on ten year bonds is currently lower than before the Fed started to lift interest rates in June 2004. America’s sovereignty over its monetary policy has therefore been eroded, with a given rise in short-term rates producing much less monetary tightening than in the past. To that extent, global monetary policy is increasingly being set in Beijing as well as in Washington.”³ While there was much truth in this statement, we think it was off the mark—monetary policy is still set in Washington. The “conundrum” had as much to do with alterations in the monetary transmission mechanism than with globalization. The rules of the game were changing and inflation volatility rose as a result.

Alteration 2: A New Transmission Channel
The debate over the relative influence of short- versus long-dated yields may distract from a more complex challenge. A new form of the bank lending channel of monetary transmission appears to be attenuating control of central banks over economic activity and liquidity conditions.⁴ More concerning, the level of control is not static, but oscillates with the business cycle.⁵

The bank lending channel, or bank capital channel in its newer form, focuses on the quantity of loans rather than the price of loans as an indicator of monetary accommodation. Inspired in the 1950s by the observation that economic activity appeared unreasonably sensitive to small changes in interest rates (the price of loans), it was conjectured that there must be a shift in the quantity of loans supplied that the pricing structure failed to detect. In effect, an entire cohort of borrowers gained or lost access to lending markets as monetary conditions shifted, amplifying the business cycle.

Small regional businesses or subprime borrowers are two cohorts that fit this description. Subprime borrowers, for example, who had been excluded from home borrowing until the early 2000s, gained unrestricted access as monetary conditions loosened around 2002, only to lose access completely as monetary conditions tightened with the onset of the financial crisis. This followed the rise and fall of leverage within the market-based shadow banking system. The price of loans, whether measured by three-month LIBOR, the size of repo haircuts or the conventional 30-year mortgage yield, would not have accurately captured monetary conditions. The bank lending channel renders pricing variables incomplete. Interest rates simply do not contain sufficient information anymore to optimally set policy.

We believe it is time to reintroduce monetary aggregates back into the policy reaction function alongside pricing variables. Central banks began abandoning quantity variables in the early 1980s as financial innovation destroyed the relationship between nominal GDP growth and all official measures of the money supply.⁶ The European Central Bank is the only major central bank to retain even a minimal consideration of the money supply today. We do not mean that traditional money supply quantities—M1, M2 and M3—should be reconsidered. Rather, new measures of money supply, including money created endogenously by the private banking system itself, should be considered. For example, Shin and Tobias (2008) found that the size of the market-based financial system’s balance sheet gives a much purer signal of monetary conditions and appears to signal changes in future economic activity quite accurately.⁷ Jonathan Wilmot of Credit Suisse made an early attempt at estimating what he calls the “shadow money stock”—the medium of exchange endogenously created within the private sector banking system (Exhibit 4). Wilmot’s data clearly demonstrate that the private shadow money stock collapsed in late 2008, but was mostly offset by a Fed-engineered increase in the public shadow money stock. Change in the traditional money stock was of little consequence. He readily admits that this information is not easily available. In fact, no official data exists on the size of the repo market or collateral usage, or even on the overall size of the securitization market. Central bankers have no long time series from which to estimate coefficients to calibrate their reaction functions. Without even the crudest of measures, monetary theorists “would almost certainly fail to adequately account for the true transmission mechanism of policy, fatally underestimate the cyclical volatility of the economy, and give no explanation at all for the repeated asset market bubbles and crashes.”⁸

In the meantime, central bankers will have to make do with traditional policy instruments, as well as unorthodox instruments hastily created during the recent financial crisis, and hope incoming shocks from the global economy are not too disruptive. The evaporation of lending to small businesses and subprime borrowers, as well as quantitative easing policies in 2008 may be an extreme example of an increasingly common phenomenon over coming business cycles.

Conclusion and Investment Thesis
Milton Friedman’s monetarist mantra that inflation always and everywhere is a monetary phenomenon remains as valid as ever. We understand that it leads to much concern over the inflationary consequences of quantitative easing policies. But on closer inspection, we feel the evidence suggests ongoing disinflationary pressures. Jonathan Wilmot’s estimation of the “shadow money stock” is a commendable first attempt, and suggests that central banks have done enough to prevent the threat of serious deflation. But far more needs to be done before central bankers will have anything reliable on which to base policy. The essential lesson is that one should not be overly confident one way or the other. Uncertainty is unusually high, and as this paper attempted to illustrate, the risks of both inflation and deflation have increased. The global context within which advanced economies operate can be choppy at times, and central bankers will have to face this turbulence with monetary policy instruments of uncertain effectiveness. All we can do is hope that inflation expectations remain anchored during this phase. Inflation insurance can be an attractive hedge in such an environment.

We included inflation-linked debt in our U.S. broad market portfolios at the market’s inception in 1997, and in European portfolios when France first issued in 1998. We have long believed that the efficiency of returns could be enhanced and that fixed-income portfolios could take a step closer to the efficient frontier with a small allocation, funded primarily from nominal government debt. The original rationale for holding inflation-linked debt was based on the diversification benefits and linkers’ ability to hedge inflation uncertainty. This rationale has strengthened dramatically over the intervening years.

Footnotes

1. This is precisely why central bank credibility and clear communication are so important, so that long real rates accurately reflect the future path of the short nominal policy rate, or else central banks lose control of the economy.
   
2. It should be noted that LIBOR is a rate with many shortcomings. It is not determined by the market but subject to monopolistic control, set behind closed doors by a collusion of large banks in London. Actual haircuts for the broader global financial system, which LIBOR is supposed to represent, may fluctuate wildly around LIBOR but, sadly, this phenomenon is largely unobservable. Thus, we expect there is far more volatility in actuality than what is captured by LIBOR. In other words, there is a natural amplification embedded in its determination which can magnify the problem. Just recently the BBA announced changes to the way in which LIBOR is set following protests as early as August 2007 and claims that it is invalid and grossly misrepresents the actual price of funds.
   
3. Warnock and Warnock (2005) corroborates, estimating that U.S. 10-year yields were about 95 basis points lower than they would have been had foreign capital inflows to the U.S. remained at their average levels.
   
4. Estrella, Arturo, “Securitization and the Efficacy of Monetary Policy”, Federal Reserve Bank of New York Economic Policy Review, Volume 8, Number 1, May 2002; and Aguiar, Alvaro & Drumond, Ines, “Monetary Policy Amplification Effects through a Bank Capital Channel”, February 2006.
   
5. “Securitization and the Bank Lending Channel”, ECB Working Paper Series No 838, December 2007; “Securitisation activity has also strengthened the capacity of banks to supply new loans to households and firms for a given amount of funding. However, we show that this capacity changes over time due to business cycle conditions: it is maximised during economic expansion when there is probably little uncertainty (in terms of Frank Knight) among investors about the valuation of structured products.”
   
6. As former Governor Bouey of the Bank of Canada put it, “we didn’t abandon the monetary aggregates, they abandoned us.”
   
7. In particular, housing investment and durable goods consumption is highly determined by the size of bank balance sheets, according to Shin and Tobias (2008), “Financial Intermediaries, Financial Stability and Monetary Policy”.
   
8. Credit Suisse, Jonothan Wilmot, ”Long Shadows, Collateral Money, Asset Bubbles, and Inflation”, May 2009.