• Monetary policy has kept interest rates too low for too long and thereby stoked the asset price bubble that burst with such devastating effect.
• Central banks focused exclusively on a very narrow definition of inflation, the increase in prices of consumer goods, thereby neglecting asset price inflation. Worse, this neglect was intentional: the central bank could always mop up after a bubble had burst, so the orthodoxy went, rather than leaning against the wind (see, for example, then-FOMC member Mishkin's speech, Enterprise Risk Management and Mortgage Lending, at the Forecasters' Club of New York, January 17, 2007).
These are both heavy charges. According to the IMF, around 30 central banks currently target inflation. Does the crisis mean that IT should be scrapped? If so, what regime should take its place?
Most leading central bankers deny the first charge. Speaking for the defence, Fed Chairman Bernanke cites a number of factors that contributed to the house price bubble, monetary policy not being among them (see Monetary Policy and the Housing Bubble at the Annual Meeting of the American Economic Association, Atlanta, Georgia, January 3, 2010). While we don't agree that monetary policy was innocent, we think the prosecution may indeed have to concede that the crisis had many causes, and that loose monetary policy was only one among many factors that contributed to the bubble - perhaps not even the most important one. But even if policymaking were to be found guilty of the first charge, IT as a policy framework would not.
The distinction between the policy framework and how policy is conducted within a given framework is not academic. Central banks keeping interest rates too low for too long may have been an error of policymaking - it does not necessarily mean that the policy framework, inflation targeting, was flawed. IT as it is practised today means that (some definition of) price stability is the medium-term goal. Central banks do have a considerable amount of leeway in achieving this goal in terms of how they set interest rates at any given point in time. Fed Chairman Bernanke calls the IT policy framework one of ‘constrained discretion': it combines commitment (to keep inflation low) with flexibility in setting policy. Hence, even if policymakers were guilty as charged in this case, the policy framework would escape sentencing. Indeed, one of the main witnesses for the prosecution, John B. Taylor, accuses policymakers of not following his eponymous Rule - effectively, of being insufficiently faithful to IT (see John B. Taylor, Housing and Monetary Policy, NBER Working Paper 13682).
What about the second accusation? Here, we think central banks are guilty as charged. With the benefit of hindsight, advocates of the ‘mopping up' approach have also changed their mind. To quote Chairman Bernanke (see again Monetary Policy and the Housing Bubble): "...we must remain open to using monetary policy as a supplementary tool for addressing [financial] risks - proceeding cautiously and always keeping in mind the inherent difficulties of that approach". However, before sentencing IT, it is worth asking what the alternatives look like.
1. Keep IT, Raise the Inflation Target
The idea of increasing the inflation target (to 4%) was recently proposed by Olivier Blanchard (see Rethinking Macroeconomic Policy, IMF Staff Position Note, Olivier Blanchard, Giovanni Dell' Ariccia, Paolo Mauro, February 12, 2010), the IMF's chief economist (most developed world central banks currently target inflation rates in the area of 2-3% though EM CBs sometimes have targets that are considerably higher). The reason is simple: the zero lower bound for nominal interest rates matters. Had inflation, and hence interest rates, been higher on average, there would have been more room to cut - hence less need to rely on fiscal policy and therefore smaller budget deficits and less debt. Again, a communication problem arises. As John Taylor says, there is something intuitive about low inflation. The furious reaction to Blanchard's proposal from the central banking community suggests little appetite for experimentation with the low inflation objective. Note that this proposal does not deal explicitly with asset prices; it merely provides central banks with more space to cut in the case of an adverse shock - from asset prices or elsewhere.
2. Price Level Targeting
Price level targeting (PT), we think, offers a number of advantages over IT (see "From Inflation Targeting to Price Level Targeting?" The Global Monetary Analyst, July 15, 2009). The most important one in the present context is the flexibility to ‘lean against the wind' of asset price misalignments without de-anchoring inflation expectations in case there is a short-term trade-off between these two objectives. However, many central bankers are hostile to PT as it would be difficult to sell to the public: everyone understands what inflation is and why keeping it low is a sensible policy objective; much fewer people will have time for a central bank that targets (a path for) the level of an index.
3. Nominal GDP Targeting
Although the idea of nominal GDP targeting (NGT) has been around for a long time - and is analytically similar to IT - it does not seem to have caught on for central banks. The main practical stumbling block seems to be that (nominal) GDP data are only available infrequently (quarterly, in most countries) and are often substantially revised afterwards. (We find this a minor issue since monetary policy as it is currently practised already uses information on GDP heavily.) A more important issue is that, taken literally, NGT implies equal weights on real GDP and inflation: if inflation is 1% higher than desired, real GDP growth would have to be pushed lower by 1%. This seems to us more rigid than current central bank practice. Note also that it does not deal with asset prices; while there is a long-run relationship between nominal GDP and asset prices, it is not clear whether the relationship is close enough at policy-relevant time horizons for NGT to provide synergies.
4. Incorporate Asset Prices
This can be done in a number of ways (see also "The Morning After", The Global Monetary Analyst, April 1, 2009).
• Include asset prices in the price index: Given the reasonably strong relationship between some asset prices and the monetary policy objectives of inflation and output, there is certainly a case for incorporating them into existing price indices. In practice, however, this approach runs into a number of difficulties, most notably what asset prices to include and what weight they should be given in the price index. However, none of these problems are insurmountable. House prices are not too volatile to render the price index useless as a target, they exhibit a stable relationship to the main objectives, and are already - imperfectly - incorporated in some price indices. The US CPI, for example, includes owner-equivalent rents (OER), but the Fed pays close attention to core PCE, rather than CPI, inflation - exactly because it gives a smaller weight to OER! The recent crisis demonstrates that academics and central bankers (and statisticians) may have missed an opportunity.
• Leaning against the wind: An important distinction in this context is whether the central bank should target a specific level or rate of growth of an asset price, or whether it should merely ‘lean against the wind'. Few, in fact, would argue for outright targeting of asset prices, while leaning against the wind has many adherents. Note that, because asset prices contain information on future demand and inflation, an inflation (forecast) targeting central bank will take them into account automatically, if indirectly. The question is therefore whether asset prices should play a role in interest rate setting over and above what they imply for output and inflation. The lesson from the recent crisis - when inflation and output were stable while house prices surged - is that the answer should be ‘yes'.
• Incorporating money and credit aggregates in monetary policy strategy: Several central banks have in the past targeted the rate of growth of (some measure of) the money supply. The Bundesbank and the Swiss National Bank did so consistently, while the Fed and the Bank of England experimented with monetary targeting for a while in the 1980s. The latter two central banks' ‘monetarist experiment' was soon abandoned - mainly because the statistical relationships between money supply growth and crucial macro variables collapsed, rendering the former useless as an intermediate target. Would a money-targeting rule have prevented the bubble? This is unclear. The growth in money supply is not tightly related to asset prices. The ECB's ‘second pillar' or ‘monetary analysis', which places emphasis on money and credit growth, is a more recent example of such a strategy. In practice, however, ‘monetary' and ‘economic' analysis - forecasts of output and inflation - usually yield the same results, and when they don't economic analysis receives greater weight. The ECB attitude demonstrates a more general issue: the implications of ‘paying attention to' money or credit for interest rate-setting are not obvious, not least because it is unclear how much independent information for output and inflation these magnitudes contain. International experience suggests that the practical importance for monetary policy decisions is probably small.
To conclude, asset price developments certainly merit greater attention by central banks than they have so far received. The recent crisis seems to advocate ‘leaning against the wind' in particular. Yet, up to a point, the advocates of ‘mopping up' made an argument with some merit: the short-term interest rate, the main central bank policy tool, is a very blunt instrument. While too little interest rate action may be ineffective to stem asset price misalignments, too much may be detrimental to the main monetary policy objectives. It could also lead to an erosion of central bank credibility if the pursuit of asset prices ends up impairing the transparency of the policy regime.
To summarise, the existing alternatives to IT either offer no (superior) treatment of asset prices, or run into severe practical difficulties. How to cut this Gordian knot? Is there a way of having one's stable inflation cake while eating one's asset price target? The answer, policymakers hope, is yes - by giving monetary policy an additional instrument to pursue the additional target with. That is, maintain IT as the framework for the main policy objectives of output and inflation, while adding another tool into the policymaker's box - a so-called ‘macroprudential rule' which links asset price developments to capital and liquidity requirements for financial intermediaries.
So even though the jury is still out, it looks increasingly likely that IT will be released on bail, with a macroprudential rule as parole officer - or as just another partner in crime? The next crisis will tell. In the meantime, both inflation and asset prices will remain under close supervision. Stay tuned.
Mexico's recent downturn in annual inflation, combined with mounting fears about the strength of the US economy, have led to growing calls that the central bank will resume lowering interest rates after staying on hold since July 2009. After all, annual inflation in 2Q averaged just 3.96%, massively undershooting the central bank's forecast range of 4.50-5.00%, while core inflation dipped in June to 3.94%, the lowest rate in nearly three years. Meanwhile, mixed incoming data from the US have sparked concerns about a potential double-dip recession which, if it materializes, would severely hurt the prospects for Mexico's economy (see "US Economics: Growth Scare", This Week in Latin America, July 12, 2010). Moreover, recent market action - Mexican interest rates have rallied across the curve - has added momentum to the growing ranks of Mexico watchers calling for a new round of interest rate cuts.
Despite low inflation figures and mixed US data of late, we find calls for new rate cuts both quite premature and simplistic, reflecting a misunderstanding of the central bank's decision-making process. A close examination of the key drivers for monetary policy suggests that, in our view, Banco de Mexico is likely to stay on hold for the rest of 2010 and into 2011. First, annual inflation likely bottomed in 2Q10 and should start moving higher as early as July. Inflation expectations, importantly, have remained stubbornly high despite the recent bout of disinflation. Second, absent a severe dislocation in the US, the Mexican economy seems to have sufficient momentum to at least match the central bank's conservative base scenario for GDP growth in this year and next. Third, the massive rally in local rates means that the market is already doing some of the easing job for Banco de Mexico. And, last, other indicators relevant to the central bank's decision-making process such as the level of unemployment, credit growth and the trajectory of the peso still suggest that Banco de Mexico's next move is more likely to be up rather than down.
Inflation: Heading Higher
Mexico watchers arguing for rate cuts point to the recent downturn in inflation; however, this downtrend in prices reflects a series of temporary factors, which pushed inflation to 3.96% on average in 2Q, undershooting the lower end of the central bank's forecast (4.50%) by over half a percentage point. Indeed, in its July 16 policy statement, the central bank pointed out that "an important part" of the downturn in inflation had been driven by "seasonal and temporary effects". As these temporary factors start to fade during the second half of the year, inflation is set to resume its upward trajectory.
Plummeting produce quotes have played a key role in pushing inflation lower in 2Q, particularly the very volatile jitomates. Indeed, jitomates and green tomatoes alone - which jointly account for just 0.60% of the CPI basket - have had a disproportionate contribution in the recent downshift in total inflation. Historically, prices of jitomates - which as of June stood nearly 20% below their average of the past ten years - have exhibited swings that were both significant and temporary in nature. And it isn't just jitomates but food in general that have helped push inflation lower: whereas overall inflation peaked most recently in March at 4.97% and moved to 3.69% by June, inflation excluding food remained broadly stable, easing marginally to 4.28% in June from 4.34% in March. And unless the recent decline in produce quotes leads to a broader drop in prices and expectations, Banco de Mexico is unlikely - and shouldn't, in our view - react by lowering interest rates.
Base effects have also helped to keep annual inflation down in recent months, concentrated mainly in prices of merchandise excluding food. Courtesy of the sharp devaluation in the exchange rate in late 2008 and early 2009, prices of goods ex-food - which include tradeable items like clothing, electronics and personal care products - did not peak until July 2009 at 6.12%, the highest level since early 2001, despite Mexico's deep slump in consumption experienced last year (see "Mexico: Reassessing the Balance of Risks", This Week in Latin America, April 20, 2009). Since then, prices of goods ex-food have been moving steadily lower to 3.83% by June, reflecting a high comparison base as well as the stability in the exchange rate.
While trailing inflation has shifted lower, inflation expectations have remained stubbornly high, staying well above the central bank's 3.0% target. Not surprisingly, In light of the benign inflation readings of late, 2010 expectations have been revised down sharply from as high as 5.28% in March to 4.66% in June, according to the central bank's monthly survey. By contrast, over the same period inflation expectations for 2011 (3.92% in June) have been stable at levels just shy of the 4.0% mark, while medium and long-term expectations have been stuck above 3.7% and 3.5%, respectively.
Time and again, Banco de Mexico has highlighted the importance of medium and long-term expectations for its policymaking process. Accordingly, we find it odd to hear more and more calls among Mexico watchers for eventual interest rate cuts even though, based on persistently high expectations for 2011 and beyond, they seem to rightfully consider the current disinflationary bout as temporary.
Economic Recovery on Track
Even if the US economy stalls during 2H10, Mexico's expansion is likely to fall within the central bank's conservative base scenario of GDP growth of at least 4.0% this year, in our view. Accordingly, we find little merit in the view that, facing mixed data from the US in recent months, Banco de Mexico will react by slashing rates. Contrary to the arguments of the rate-cut camp, in fact, the July 16 policy statement contained a more constructive tone about the prospects for the Mexican economy after the central bank acknowledged the strength in industrial output while consumption and investment, though still somewhat depressed, "appeared to be already showing a favorable shift in trend". By contrast, in June the central bank limited itself to pointing out that consumption and investment "had recently displayed some improvement".
Given the ongoing V-shaped recovery in industrial output and signs that recovery is starting to broaden to the consumer side as well, reaching at least 4.0% GDP growth in 2010 seems like a very low threshold, in our view (see "Mexico: More than Just Cyclical?" and "Mexico: Consumers - Waking Up", This Week in Latin America, June 14 and June 21, respectively). In 1Q10, the economy rose 4.3% from a year earlier and the strength of recent economic releases - such as the 7.2% jump in April's GDP proxy (IGAE) and 8.4% in May industrial output - suggests that GDP growth in 2Q could top the 7% mark. Against this backdrop, even assuming very modest or even no sequential growth at all in 2H, the Mexican economy would comfortably top 4.0% growth this year.
Whereas overall economic activity is showing good momentum, fixed investment has lagged significantly. Since bottoming in May 2009, the economy has expanded almost 9% as measured by the monthly GDP proxy IGAE. Meanwhile, in April 2010 the indicator of fixed investment stood barely 3% above the May 2009 lows. The upshot has been rising capacity utilization, which is currently running just one percentage point below its long-term average, according to our calculations. With economic activity on the rise and investment still sluggish, we suspect that the central bank's base case scenario for the output gap to close at some point during 2011 is still valid, leaving little room for easing policy rates.
Other indicators relevant to the outlook for monetary policy still suggest that Banco de Mexico's next move is more likely to be tightening, in our view (see "Mexico: The Case for the Doves", This Week in Latin America, April 5, 2010). Total credit is already on the rise, expanding nearly 6% annualized in the three months ending May, according to our calculations. The slump in consumer credit - which was the hardest-hit credit segment - is clearly subsiding after contracting about 4% annualized over the same period, a far cry from the double-digit sequential declines observed in late 2008 and during most of 2009. The economy has created jobs uninterruptedly since last August, recovering by June 2010 all the formal positions lost during the recession. And while the rate of unemployment has moved slightly higher in the past couple months, this apparent deterioration has been in part caused by a surge in labor participation.
A Word on Monetary Conditions
With inflation set to rise in coming months, this should translate into an easing in monetary conditions - without requiring the central bank to act - as real interest rates move lower. Our proprietary Monetary Conditions Index, in fact, has been relatively stable in recent months as the modest increase in real rates - caused by the decline in inflation - has been in part offset by a weaker exchange rate. Assuming no change in the exchange rate, then rising inflation should lead to a more expansionary policy stance. And insofar as the central bank focuses not just on overnight rates but also looks at longer rates to assess broader monetary conditions, then the massive rally in local rates means that the market is already doing some of the easing job for Banco de Mexico.
Bottom Line
Despite recent low inflation readings in Mexico and mounting fears of a double-dip recession in the US, Mexico's central bank is unlikely to cut interest rates anytime soon. Not only has the downturn in inflation been caused by a series of temporary factors which should fade over the course of the year, but expectations have remained stubbornly high. And even if the US stalls in coming quarters, the Mexican economy seems to have sufficient momentum to at least match the central bank's conservative base scenario for GDP growth. Against this backdrop, the growing ranks of Mexico watchers looking for imminent interest rate cuts are likely to be disappointed as the factors shaping the central bank's decision-making process still suggest that Banco de Mexico's next move is more likely to be up rather than down.
Read Full Article »