An Unprecedented $660 Billion In Excess Debt Demand, And What It Means For Bond Yields
When the BOJ announced two weeks ago the full details of its expanded easing program, which amounts to monetizing a whopping $720 billion in government bonds over the next year (a move which makes even the Fed's own open-ended QE appear like child's play in perspective), one thing it did was lay to rest any hope of a rotation, great or non-great, out of bonds and into equities. The reason is simple: while the Fed is en route to monetize $1,080 billion in UST and MBS debt in the current year, when there is just $760 billion in net US issuance, what the BOJ has done is add a bid for another $720 billion when Japanese net supply of debt is just $320 billion in the next 12 months. In other words, between Japan and the US, there is now some $660 billion in secondary market debt that the two banks will have to purchase over and above what their respective treasury departments will issue.
This means that the two biggest central banks are about to be perfectly price ambivalent as to what cost the monetize nearly two thirds of a trillion in debt: their mandate is to expand their assets at any cost, which means buying up debt from the secondary market at any price. The immediate consequence is that the best performing trade of 2011 and 2012 - frontrunning the Fed's purchase of the long end - is about to come back with a vengeance, as speculators buy up every piece of duration debt that is not nailed down, knowing full well they will be able to sell it right back to Bernanke and Kuroda in the future at whatever price they so desire.
The chart summarizing the massive disconnect between the monetization mandate and the net treasury issuance in the US and Japan, and showing the $660 billion delta, is below:
This is how JPM explains how much further yields will likely fall as a result:
We mentioned last week the displacement caused by BoJ purchases this year. The offset that central bank purchases provide to government debt issuance, represent the so called Demand or Flow Effect of QE. In fact, central bank purchases more than offset government bond supply this year. The BoJ is going to surpass government debt supply this year by buying $720bn of government debt vs. government net supply of $320bn, resulting in a net withdrawal of $400bn of government debt securities from bond markets. Assuming $85bn per month purchases for the whole year the Fed looks set to buy $1020bn of USTs and MBS securities this year vs. net issuance of $760bn, so a net withdrawal of $260bn from bond markets.
But there is another impact caused by BoJ purchases. The stock of excess reserves is likely set to rise sharply this year roughly in line with the amount of BoJ purchases, i.e. by $720bn, at the same time as the stock of government debt held outside the BoJ shrinks by $400bn. Similarly, the Fed looks set to inject close to $1000bn of excess reserves into the banking system at the same time as it withdraws $260bn of government securities.
This additional effect caused by the rise in central bank reserves, represents the Liquidity or Stock Effect of QE. We discussed this stock effect in a previous F&L on Dec 14th. The idea is that QE creates scarcity by making one form of collateral (government bonds) more expensive relative to another (zero yielding reserves). Given that the banking system cannot get rid of reserves, these zero yielding reserves, become the “hot potato” that banks and other investors try to pass to each other until the relative pricing is adjusted enough to remove the incentive for banks or investors to get rid of these reserves.
And because the relative scarcity of these two forms of collateral depends on relative stocks rather than flows, the price effect would remain even in an environment where flows disappear. That is, even if central banks were to stop purchasing bonds, making the flow or demand effect of QE to go away, the liquidity or stock effect would continue to affect the relative pricing of government bonds vs. zero yielding reserves.
By how much is this Liquidity or Stock effect of QE affected by the BoJ policy? The change in the relative stocks of these two forms of collateral, excess reserves vs. the government debt held outside the central banks, in the case of BoJ at ($720bn - (-$400bn)) = $1120bn, comparable to that caused by the Fed ($1020bn - (-260bn))=$1280bn.
As a result, the liquidity or stock effect looks set to intensify this year. To quantify this liquidity effect across the G4 we use the ratio of excess reserves at the Fed/BoJ/ECB/BoE, divided by the stock of government securities held outside these central banks. Factoring additional QE by the Fed, BoJ, and BoE (we expect an extra £50bn of Gilt purchases by the BoE this year) and another reduction of around €200bn of excess reserves in the Euro area banking system due to further LTRO repayments, we project that this liquidity ratio will rise from 10% at the end of March to around 14% by year end. Again, under the assumption that bond buying by the Fed will continue at $85bn per month until the end of the year.
What does this mean for government bond yields? One simple way is to look at the changes in this liquidity ratio vs. changes in government bond yields. This is shown in Figure 1. Figure 1 shows there were two major episodes in the evolution of the liquidity ratio in recent years. The first major episode was in H2 2008, immediately after the Lehman crisis, when the liquidity ratio jumped from 0.5% to 5.0%. The second major episode was during 2011 and the first half of 2012. During these 6 quarters, the liquidity ratio rose from 5.5% to 11.0%.
Figure 1 shows these two episodes saw the biggest declines in bond yields over the past five years. During the first episode the government bond yield of our GBI Global index declined by 120bp. During the second episode the GBI Global index bond yield declined by 70bp. Given that the two episodes were accompanied by a similar 5% increase in the liquidity ratio, this suggests there are diminishing effects to QE, i.e. each further 5% increase in the liquidity ratio exerts smaller downward pressure on bond yields. It would be thus reasonable to assume that another episode of a 5% rise in the liquidity ratio for example should lower bonds yields by less than 70bp, perhaps by around 40bp or so.
With the caveat that it is difficult to separate the flow, stock or signaling effects of QE, mechanically this simplistic exercise suggests that a 4% increase in the liquidity ratio over the next three quarters could push bond yields lower by around 30bp (from March end levels of 1.75% for the GBI Global index yield).
To summarize: in the coming months, bond yields will continue to slide as more speculators realize that they can extract any price from not one but two central banks, now desperate to buy not only all net primary issuance but a massive $660 billion in secondary market UST and JGB demand. This means that with central bank balance sheets becoming ever more ridiculously large, a phenomenon which even the most clueless textbook economists will admit will ultimately result in runaway inflation, the yields on the primary instrument which in normal times reflects the threat of runaway inflation, the long bond, will continue declining, and signaling to an ever dumber and more centrally-planned market that there is not one iota of inflation on the horizon. Taking the thought experiment to its ludicrous end, we can visualize the Fed and the BOJ's balance sheets hitting infinity as nominal yields become negative.
Thought experiments aside, unless of course the Fed and the BOJ are determined on becoming the only bidders in both the primary and secondary market, this means that the weak hands will likely continue to puke their gold holdings. However, and much to the chagrin of SocGen, gold is and has always been an inflation hedge, and we, for one, are delighted to take every opportunity to rotate out of ever more diluted fiat paper into hard currencies (and assets). The reason is that inevitably the moment when the central bank wheels hit the road will come, and it will be not so much a question of imminent price reaction, but what the long-bond, which will suddenly find itself without central bank bidders, and thus ostensibly bidless, does. What it will do is collapse, sending long-yields exploding, and making up for years of faulty inflationary signalling. Ironically this may come at a time when the central bank balance sheets are indeed contracting. However, without them present as bidders of last resort, watch out below if one is long the 30 year.
It is at that point that all those long bond to gold algos will wake up, and finally see the event horizon which for four years had been hidden by the global central banks' coordinated attempt at centrally-planning all asset levels.