Why central banks should buy in the worst way possible
Since the August 4 announcement that the Bank of England will purchase corporate bonds again, much chin stroking has occurred about the differences between the Bank’s approach and that of the European Central Bank. In at least one respect, though, the BoE's scheme seems superior.
What’s the point of quantitative easing? At this late stage, with talk of helicopter money in the wings and negative rates the norm rather than the exception, it’s hard to be sure. At times, it feels like QE is an end in itself.
Central banks need to do QE because that’s what central banks do nowadays, and they can’t stand idly by while economies sputter. Or they do QE because markets have grown to find it reassuring and comforting. It’s supposed to encourage lending, or lower the cost of lending, or expand the monetary base, or force investors into riskier real economy lending, or spur new investment projects, or lower the debt burden on households and governments. Maybe all of the above.
But, while all of these aims are sort of blurred together, they have a major effect on how central banks should actually conduct purchases.
A central bank can try to buy assets below market rate (like a normal investor), at a ‘market neutral’ rate (on the absurd theory that you can reallocate hundreds of billions of euros without causing price distortions), slightly above market (acknowledging that you have to pay up to take size), or way above market (to have the maximum possible effect on yields).
The first approach is a non-starter. No central bank will be able to do any sizeable programme of asset purchases and find bargains. Central banks intentionally signal their willingness to ease monetary policy ahead of time, and ensure that the whole fixed income market gets the chance to front run any actual purchases.
The second and third approaches seem most typical of the ECB’s approach. When the ECB plays in a new issue, it always tries to follow, rather than lead, on pricing. Like an elephant tiptoeing in ballet shoes, it will sneak in for perhaps 40% of an issue without doing any price anchoring — though having reliable orders in size might give the lead managers the chance to be punchier on price.
In the secondary market, it’s closer to the third approach. The ECB has more or less demonstrated that the lack of offer-side liquidity in corporate and covered bonds is a myth — if you’re willing to be aggressive enough on price, it’s possible to find bonds. Most private investors would rightly fight their dealers over every last cent, but the ECB has more to fear from missing its monthly targets than it does from overpaying for assets.
Arguably, though, it should be intentionally throwing its weight around. If the purpose of the purchases is to lower debt costs, encourage financing of riskier assets and flatten the yield curve, it’s best to do that by purchasing with as much price impact as possible. Ditch the idea that it’s possible, or desirable, to create billions in new euros or sterling without wrenching distortions on asset prices.
Acting on the price, rather than the quantity, of money doesn’t exactly chime with monetary orthodoxy — but we’re so far through the looking glass on what counts as monetary orthodoxy it hardly matters. Intentionally purchasing to maximise price impact means central banks keep more ammunition left over if they need it, and cause fewer technical distortions in markets.
All other things being equal, if central banks achieve the same price change with smaller quantities of new money, that means fewer bonds locked up with the monetary authorities. There will still be negative yields for some assets, but other markets will be less technically squeezed, with better functioning repo markets and more fluid trading. There’s no point a central bank buying totally off-market — it needs to shift the prevailing market price level in line with its purchases — but it can try to maximise its impact.
Which brings us to the Bank of England. At the moment, the conditions of the Bank’s new corporate bond buying programme require an issue to have been a month outstanding before it can buy.
That’s exactly when it starts to be truly illiquid. Most new issues trade heavily in the first few days after issuance, and settle down into buy-and-hold hands thereafter. Data from the International Capital Market Association’s corporate bond market study shows that from day six after issuance, most new issues trade less than twice a day. From day 30, it’s closer to once a day.
For now, then, the Bank is maximising its impact. £10bn of purchases will clearly have more impact in an environment where average turnover is £1m per day, in one ticket, than if it’s burned away in big primary purchases.
Though it is considering primary buying, it should be congratulated for resisting. Though it will be trying to ramp up buying in an illiquid market, that’s a feature, not a bug.