Since the ECB launched its sovereign debt QE, initially known as PSPP, in March 2015 and later expanded to include corporate debt, or CSPP, in June 2016, the world's biggest hedge fund central bank has created enough money out of thin air to purchase bonds with no consideration for price to grow its balance sheet, i.e. investment portfolio, by €1.89 trillion.
Meanwhile, over the entire QE period, net European bond new issuance has only amounted to €394 billion - only one-fifth of what the ECB has bought - and that only after picking up recently. In fact, through much of 2016, there was hardly any net issuance at all according to Citi data.
Here, as Citi notes, It’s hardly rocket science that for every bond the ECB has bought there must have been a seller – either a new issuer or an existing holder, which means:
Net € FI issuance = Domestic net buying of € FI + Foreign net buying of € FI + ECB net buying of € FI
Imbalances between desired issuance and desired holdings at the prevailing market price are what drive valuation changes until equilibrium is found. Put differently, if the ECB bought a bond from an investor who wished to remain in the € fixed income market, then that investor would buy from another investor, who could buy from yet another, but unless there was new issuance to invest in eventually prices would reach levels where someone would take the money out and put it somewhere else.
But who has been selling to the ECB? And where have they been putting their money?
That's the question Citi's Hans Lorenzen set out to answer, and since per the math above, net of issuance holdings of private investors must have fallen by more than €1.5 trillion, the answer would be rather material. Put that number into context, the €1.5 trillion in debt that someone sold without replacing, is equivalent to more than 9% of €-denominated bonds outstanding at the start of the program. Those are bonds which used to be held by private investors, who have now been given cash and have to park that cash somewhere else.
As Citi notes, "It truly is crowding out on an unprecedented scale."
Going back to Citi's question, here is the answer in two parts.
First, the "who" sold this €1.5 trillion in private holdings:
Using ECB data, we know that private banks have beem major sellers, to the tune of €645 billion since the start of QE, making up more than 40% of the decline in private holdings. Here the net selling has mostly been of government bonds (€293bn) other MFIs (€273bn), while corporate and other bonds only amount to €70bn. Aside from banks, Citi calculates that while non-resident European investors have sold €400bn since Q1 2015, with non-bank private Euroarea investors filling the gap of €795bn. This calculation challenges the predominant assumption that the selling to the ECB has mostly been done by foreigners, as much of the non-bank net selling must have come from other domestic investors. When one adds Eurozone banks, it is clear that the majority of private selling in aggregate has been domestic.
The chart below breaks down the transactions in bonds issued by Euro-area residents by investor type. Aside from banks, the main sellers have been households and other financial institutions.
Second, where did the money go?
While the answer will hardly come as a surprise, there are - intuitively - four destinations where a euro pulled out of € fixed income could move into.
- stay in fixed income, but move into bonds denominated in other currencies;
- move out of fixed income and into another asset class (domestic or foreign), like equities;
- move into money markets;
- leave the securities market altogether, in which case you'd expect it to show up as a deposit (with a domestic or a foreign bank
While there are some potential complications here, mostly because there is no explicit data revealing the "mirror image" for the private selling of €-denominated bonds, if one lines up the transactions against Citi's proxy, consisting of net purchases of European equities, money market flows, and non-government deposits, and the directional terms of the resulting asset disposition proceeds emerge, or as Citi summarizes, "When investors
have been selling bonds, investments in our proxy have mostly tended to rise."
Furthermore, as shown in the second chart below, splitting up these “proxy investment outlets” into their constituent parts, it becomes clear that the bulk of the “delta” from before QE in 2014 is in an increase the rate of deposit accumulation and an increase in outflows from Eurozone investments. However, since QE began, the deposit accumulation has continued, but the purchases of equities and especially foreign investments have grown, and have accelerated markedly this year.
In short: the ECB purchased €1.5 trillion in bonds, mostly from European banks and domestic investors, with no regard for prices thereby virtually assuring booked profits for the sellers, who then turned around and purchased domestic equities, foreign investments, or converted the money into deposits and money market instruments.
And so, with the ECB set to taper with trial balloons that the ECB could cut its monthly QE by half or more, what happens next now that this swap is about to be throttled by more than 50%? Will households sell more or less bonds, and what will happen to yields? We will present one answer - an answer which the central banks do not want to hear - shortly.