The latest myth of a European recovery came crashing down two weeks ago when Eurostat reported an inflation print of 0.7% (putting Europe's official inflation below that of Japan's 1.1%), followed promptly by a surprise rate cut by Mario Draghi which achieves nothing but sends a message that the ECB is, impotently, watching the collapse in European inflation and loan creation coupled by an ongoing rise in unemployment to record levels (not to mention the record prints in the amount of peripheral bad debt).
Needless to say, all of this is largely aggravated by the soaring EURUSD, which until a week ago was trading at a two year high against the dollar, and while helpful for Germany, makes the so-needed external rebelancing of the peripheral Eurozone countries next to impossible. Which means that like it or not, and certainly as long as hawkish Germany says "nein", Draghi is stuck in a corner when it comes to truly decisive inflation-boosting actions.
But what is Draghi to do? Well, according to BNP's Paul-Mortimer Lee, it should join the "no holds barred" monetary "policy" of the Fed and the BOJ, and promptly resume a €50 billion per month QE.
Why? Some preliminary perspectives from BNP:
Some economies in the eurozone are already in deflation. This has adverse effects on resource allocation (nominal rigidities in some prices mean that relative prices do not adjust enough) and, of course, sets in train adverse debt dynamics. It deters spending today, because things will be cheaper tomorrow. When significantly negative real interest rates are warranted, zero or negative inflation makes these impossible to achieve. Countries that are undergoing structural reform need lower real interest rates than they would otherwise have. Extremely low or negative inflation, therefore, militates against structural reform.
One fact sums up the parlous state of affairs: the eurozone has lower inflation than Japan (0.7% versus 1.1%).
At such low levels of inflation, deflation is a real danger. How big a threat is it? The inflation forecasting literature says that since the Great Moderation, the best way to forecast inflation over longer periods is a random walk. (Over short periods, judgemental forecasts, (eg, taking account of energy prices, tax increases etc) are preferable.) A random walk model says that decreases in inflation are as likely as increases. If we start from 0.7% inflation, the central forecast of a random walk model will be that inflation stays around its current level, ie, that deflation is less than 50% likely. Our own forecasts agree with this but also – like the ECB judges – indicate that inflation will remain at very low levels for a very long period.
Why has inflation in Europe proper not become outright deflation across the continent, especially when considering the record low rate of growth (in reality, contraction) of loans to private sector companies? The only silver lining are "inflation expectations" which are still anchored somewhere aroun the 2% area. However, that may not last.
If inflation expectations break down, then as rates are close to the zero bound, getting them back up again could be extremely difficult. In fact, it may be impossible. Don’t wait until you are drowning to think about looking for a lifejacket. That is one of the lessons of Japan, waiting until too late can leave you locked into an insoluble problem.
Why then did the hawkish members not want a rate cut? It is difficult to be sure but we suspect it is a combination of strategic and tactical considerations (our comments in parentheses):
- Inflation expectations are stable; therefore there will be a gravitational pull upward in inflation over time. The ECB’s credibility will do the job. (Only if expectations hold up, which is questionable if inflation stays around 1% or below for a prolonged period, which seems likely even on an optimistic assessment.)
- Core inflation is very stable in Germany, which shows that the lows for inflation elsewhere are cyclical not structural and as growth picks up inflation will come back to target. (Our perspective: There is no sign the output gap will close any time soon. The level of the output gap is likely to put downward pressure on inflation.)
- The disinflation in the eurozone as a whole primarily reflects relative price changes of peripheral countries. As they regain competitiveness, their inflation will converge again on the German level. (We would argue this can take a very long time – how many years will it take to get Spanish unemployment back to pre-crisis levels? Also, if the gravitational pull of German inflation was so strong, how did the periphery lose competitiveness in the first place?)
- The severity of the fall in inflation is overstated by developments in energy prices. (There is truth in this, but they were not singing from the same hymn-sheet when energy prices were rising.)
- The rate cut will not do anything. (We agree.)
- The sooner the rate cut ammunition is used up the closer we come to QE. (We agree, but see this as entirely beneficial.)
- We might get lucky and inflation will just pick up. (Much less than a 50% chance in our view.)
What all this amounts to is that the hawks will continue to resist easing until either inflation expectations crack, by which time it will be too late to recover, or until German inflation sinks to the levels other countries are experiencing today. With full employment in Germany and a minimum wage on the horizon for 2015, that looks likely only in a very severe recession. In either case, what this boils down to is that the hawks will only support radical easing in a disaster scenario, by which time it will probably be too late and deflation will be entrenched. Presumably the ECB’s minutes will be published in Japanese.
Needless to say, all of the above is only correct and valid in the confines of a Keynesian macroeconomic framework where deflation is a bete noire and must be avoided at all costs, and instead the taxation on money also known as inflation is the only saving grace. We will refrain from the well-known philosophical refrain that the only reason the world is in its current inescapable predicament is due to a faulty economic framework which has advocated precisely more of the same for about a century and continue within the false dichotomy laid out as gospel.
So assuming the ECB does have to stimulate inflation what can it do? It is here that the BNP strategist gets excited:
Does the ECB have enough conventional ammunition left to narrow the output gap by enough to get inflation from less than 1% to close to 2%? No way. According to conventional models, it will struggle to stop inflation from decelerating further.
Where should rates be? Let’s consider the Taylor rule. It’s by no means infallible as a guide to where monetary policy setting should be, but it provides some sort of benchmark. First, let’s ask where real policy rates should be in equilibrium. We estimate trend growth to be about 1%, so that is a reasonable starting point, which would suggest a nominal equilibrium rate of 1.7% given current inflation.
But we are not in equilibrium. From 1.7% we have to subtract 0.5 times the output gap. We do not know how large this is, but let’s take the OECD figure of 4%, taking our Taylor rule rate down to -0.3%. However, we haven’t finished yet because inflation is below target. To stabilise inflation, deviations of inflation from target have to be met by a larger reduction in nominal rates otherwise the real rate would not fall and the economy would not stabilise and bring inflation back up. The Taylor rule says subtract 1.5 times the inflation shortfall, which is 1% if we assume “below but close to 2%” means 1.7%, so we subtract another 1.5% points, giving us a Taylor rule appropriate rate of -1.8%.
Since this is in negative territory, only a deposit rate cut can deliver it. However, we doubt that the ECB would take the deposit rate so far into negative territory for fear of a variety of distortions and adverse effects that would result (see link here to my note on negative deposit rates earlier this year). Failing this, unorthodox monetary policies are called for, such as expanding the balance sheet.
As an aside, the Taylor rule may cast light on why the hawks do not want rate cuts. The Breugel think-tank (http://www.bruegel.org/nc/blog/detail/article/1151-15-percent-to-plus-4-... rule-interest-rates-for-euro-area-countries) recently calculated that the appropriate Taylor rule rate (based on an unemployment gap, that gives an appropriate rate higher than our estimates) in the eurozone varies between -15% (Greece) and +4% (Germany). Unfortunately it would appear some ECB Board members are not voting in the interests of the eurozone as a whole. Of course, too low a rate is a problem for Germany, for asset prices and perhaps inflation. But macro-prudential measures should be used to offset adverse effects rather than forcing others to have massively too-high rates.
Which means, according to BNP, that the only realistic deus ex machina (ignoring that the Fed has so far failed to stimulate broad CPI-defined inflation for five years running) would be, you guessed it, QE. A lot of it.
How many government bonds would the ECB need to buy to achieve its past batting average? M3 is almost EUR 10trn, meaning that 1% of M3 is EUR 100bn. Credit to general government is almost EUR 3½trn and is up just 0.7% y/y. Taking M3 up from its September y/y rate of 2.1% to the old reference rate of 4½% would require about EUR 240bn of QE in 2014. Taking M3 to its average growth rate of 5.8% would require EUR 370bn. But there is also the cumulative shortfall relative to trend to make up – some EUR 1.2trn. We would not advocate closing all this in one year, but taking this into account, we would suggest EUR 50bn a month for purchases over the first year or two.
We see the attractions of QE as being:
- Increasing the rate of monetary growth to reduce disinflationary pressures;
- Reducing long-term rates and stimulating growth;
- Being likely to reduce risk premia in the economy;
- Being likely to boost asset prices, such as stocks and credit (namely, Japan);
- Being likely to soften the EUR on the FX markets;
- Being likely to stabilise inflationary expectations, which might otherwise sink;
- Demonstrating the ECB’s commitment to price stability and to stick to its Treaty obligations; and
- Reducing the quantity of government bonds on banks’ balance sheets while increasing their liabilities (deposits due to higher M3) would stimulate private credit by reducing crowding out.
Of course, there are downsides, too:
- There would probably be legal challenges on the basis that the ECB was embarking on monetary financing of governments (though buying in the secondary market would circumvent this);
- In some countries, such as Germany, there might be an atavistic adverse reaction;
- The programme would not be self-extinguishing in the same way as LTROs were (a good thing, in our view);
- Inflation expectations may rise (again, a good thing, but challenged somewhat by the experience of the US and Japan, where QE has hardly resulted in rampant inflation);
- It could reduce the incentive for governments to carry on with fiscal consolidation because financing would be easier; and
- Asset bubbles might result from it.
- What assets would be bought?
The last question is one we believe will exercise the ECB a good deal. There is a question as to whether private debt markets would be deep and uniform enough across the eurozone to allow the scale of purchases required. When the ECB bought covered bonds, its programmes were limited in scope. (The first scheme in 2009-10 resulted in total purchases of EUR 60bn, equivalent to around 0.7% of annual eurozone GDP. The second programme in 2011-12 was smaller still, resulting in total purchases of EUR 16bn versus the initial target of EUR 40bn.)
Buying government bonds is the obvious route. However, buying only peripheral bonds would not seem to be an option. (It would, effectively, be the OMT, but without a programme and without conditionality.) So, buying across the spectrum of government bonds seems natural (with the proportions determined by the ECB’s capital key).
There would probably be objections to buying German Bunds, as:
- Rates are already low;
- Germany, and German business, has no problem financing itself; and
- Financial and monetary conditions in Germany are already easy; why make them easier still?
We would respond that QE in a European context would work in a different way to the US:
- In the US, lowering the risk-free rate lowers rates for private borrowers also;
- Private bond finance (mortgages as well as corporate debt issuance) is stimulated and this benefits the economy;
- In Europe, the private bond market is less developed and firms largely finance through banks;
- Similarly, mortgage finance in the eurozone comes much more from banks than from mortgage bonds ;
- Therefore, QE in Europe would not work in the same way as in the US;
- Much more of the effect of ECB QE would come through the exchange rate;
- It is those assets held by foreigners that the ECB should target in its purchases, encouraging a very low rate that makes the assets unattractive to current foreign holders; and
- So, Bunds and OATs and the bonds of the other core countries are precisely the assets the ECB should buy.
Over 60% of German Bunds are held by non-residents; the proportion of foreign holdings of OATs is also high. Therefore, buying these assets would not only benefit the economies of the issuer of the securities, but also the countries of the holders. The benchmark status of the Bund would lower all yields in the eurozone and need not bring about spread widening – substitution and a search for yield would be likely to narrow spreads.
Telling a Weimar-PTSD'ed Germany that the ECB is coming and will almost exclusively monetize just Germany's bonds? Good luck.
We would expect the ECB to exhaust other channels before resorting to QE – cutting the refi rate below 25bp, and maybe opting for a negative deposit rate to try to get the exchange rate down. It may engage in forward guidance more actively. However, the power of such measures looks limited. If we have further downward surprises to inflation, as we had this month, there will be very little alternative, if the ECB is not to accept a magnified risk of deflation, other than to go for QE. Inflation data and inflation expectations will be crucial in determining what the ECB does over the coming months.
It is very unlikely that the hawks will agree to such measures until disaster is already at the door, so to get the right result for the eurozone, Mr Draghi will have to risk resignations. Otherwise, he should take Japanese lessons.
Of course, BNP is ultimately correct as the European experiment, which is doomed for the simple reason that Europe will never be able to achieve the kind of internal rebalancing it needs absent standalone currencies, will require every weapon in the ECB's arsenal, and sooner or later the ECB, too, will succumb to the same monetary lunacy that has gripped the rest of the developed world in the ongoing "all in" bet to reflate or bust. All logical arguments that outright monetization of bonds are prohibited by various European charters will be ignored: after all, there is "political capital" at stake, and as Mario Draghi has made it clear there is no "Plan B."
Which means the only question is when will Europe join the lunaprint asylum: for the sake of the systemic reset we hope the answer is sooner rather than later.