The risk of the USA slipping into deflation is a lot higher than most people think. Current headline and core measures of CPI / PCE are being help up by once off increases in things like tobacco excise – and if we strip these out, underlying inflation pressures are very weak.
After Bernanke gave his helicopter speech in late 2002, core PCE bottomed at 1.4% and core CPI got down to 1.1% – these low numbers were partly due to once off factors; median CPI bottomed out at 1.7%y/y in late 2003, and trimmed mean PCE also bottomed at 1.7%y/y.
As this downturn has been very sharp, year ended measures of inflation pressure miss much of the action. Using 6mma AR measures gives a better sense of deflation pressures, but are not so sensitive as to be meaningless.
On these measures, headline CPI is 2.8%y/y, core CPI is 1.4%y/y, and median CPI is 0.4%y/y. Headline PCE is 2.2%y/y, core PCE is 1.2%y/y, and the trimmed mean PCE is 0.8%y/y. The 2003 cycle bottom for the median CPI was 1.4%y/y (100bps above current), and the cycle low for trimmed mean PCE was 1.4% (60bps above current).
As inflation falls with a lag following an increase in the rate of unemployment, it’s unlikely that inflation pressures have bottomed out just yet. To capture this lag, i’ve scatter plotted the 18m change in the rate of unemployment, from 12m ago, against the rate of inflation. I’ve marked the apparent peak in the change in the unemployment rate (October 2009), and the associated rate of inflation – if that is the peak in the unemployment rate, we might expect inflation to bottom out in late 2010.
We can do a little better than this for a forecast, however, as past inflation is a good predictor of future inflation – I’ve done that here. Showing both the simply forecast path for PCE implied by the scatter plot regression, and an alternate forecast using unemployment and past inflation. Depending on which approach we take, we might expect PCE to bottom out around 0.3%y/y or 0.4%y/y some time in late 2010.
Now it’s worth remembering what Bernanke said in 2002:
when inflation is already low and the fundamentals of the economy suddenly deteriorate, the central bank should act more preemptively and more aggressively than usual in cutting rates (Orphanides and Wieland, 2000; Reifschneider and Williams, 2000; Ahearne et al., 2002). By moving decisively and early, the Fed may be able to prevent the economy from slipping into deflation, with the special problems that entails.
If my inflation forecast is right, the Fed won’t be doing anything with the cash rate in 2010.
So what are they doing with the discount rate hike?
I think they are preparing this capability to defuse a run-away credit bomb. In the unlikely case that the machinery of policy jags, and things start to work (which basically means credit starts to grow once again), credit could explode. There are never-before seen amounts of reserves in the system. If used to their full potential, these reserves would support a never before seen level of deposit creation (lending) on the part of banks.
The Fed would have to intervene – especially as this would surely be a bubble – to halt the process, to head off a very serious asset and price inflation; and limit the damage caused by a really serious bubble.
I think that the Fed is preparing the ground for a sudden withdrawal / immobilisation of reserves, just in case this occurs.
But i don’t think they believe think it’s likely. It’s a high risk scenario, with a low probability of occurring, and the Fed is taking out some insurance – as they should. Most likely, we’ll see a limp recovery, and if that’s the case, the fed funds rate will remain at 0bps to 25bps for all of 2010, and probably 2011 as well.
If this is the case, the short end remains safe for carry trades, breakevens are way too wide, and hence the level of rates is too high right across the US curve.
Once the market comes to understand this, the fed’s having prepared for the small possibility that the reserve bomb explodes will enhance the fed’s inflation fighting cred. This should lower the inflation premium in rates markets.
Having prepared the reserve-bomb squad, should inflation turn out to be lower than expected, the Fed would have the credibility to buy more bonds. I think that this time, they’d just buy USTs. The fed used to only deal in GC, and I think that they’d like to get back there…