Everyone expects the Fed to announced the next round of QE this week.
George Goncalves at Nomura has gone back through history to help put some context on the imminent announcement:
Shock and Awe: We do not expect the Fed to go the “shock and awe” route ($500
bn plus) because it doesn’t want to pre-commit to a number that the market would
expect them to buy in full and we believe that by remaining flexible, it allows the
Fed to remain responsive to the upcoming data. That said, this past week, various
newswires held surveys where the size of QE expectations ranged from an
average of $850 bn from a Reuters poll to a WSJ median from other primary
delears that was closer to $1 tn. The obvious risk to our Fed call and rates view is
that the Fed comes in with a large program.
Headline Effect: The Fed could announce a Treasury buy program of $1
trillion,which for example, could be broken down into two pieces. If it makes
assumptions on the future path of MBS prepays, the Fed could state something like,
“the Fed has scope to purchase up to $1 tn in USTs via new purchases and rolling
over of MBS” (assuming $450-500 bn from prepays/Agency debt paydowns and
$500 bn from reserves). In our opinion, using this approach could be rationalized
to elicit a headline effect from a large number.
Smoke and Mirrors: Committing to buying Treasuries for an “extended period of
time” without a hard final total number is another extreme option. In the 1940s
through the 1951 Treasury-Fed Accord, the Fed kept long-term rates for bonds at
around 2.5% to help fund the war time period deficits, but also help immunize bond
losses to the banking system. Although written about from an exit strategy
perspective, Aichi Amemiya, US economist wrote about this time period in a report
last year – see link. Mr. Amemiya highlighted in his Figure 4 chart that interest
rates did not budge, even though during that Fed operation, there were periods of
high inflation and deflation.
We highlight this 1941-51 period, known as the pre-Accord period, because the
Fed did not have an explicit target of how many Treasuries it purchased. Instead, it
maintained a yield target (see link page 9 – for more information from a Fed report
published in the 1990s). Some argue that using this approach, the Fed might not
need to buy as many securities as it would in a pre-defined LSAP (large scale
asset purchase) since the Fed would be always on the bid side at a certain level.
This approach would be the hardest to implement, but there is historical precedent
for such a policy and it did manage to keep rates in a tight range (Figure 2).
We think this sort of policy would be too dramatic at this point because it comes
with serious unintended consequences. The bond market now is highly dependent
on foreign buying and the USD is the reserve currency. Dissenfranching global
investors might make the exit strategy many years down the road very hard to
execute should the Fed decided to yield target.
The accompanying chart is very useful, and should probably be saved away for future reference: