By now everyone is probably familiar with one of the scariest financial charts created recently by Bank of America: it shows that not only have central banks injected a record $15.1 trillion in liquidity since the crisis, but in 2017 alone - a time when the global economy is supposedly improving - they added a record $1.5 trillion, or as BofA's Michael Hartnett calculated, $3.1 trillion annualized.
Or maybe not: in a little noticed comment from Hartnett made later in June, Hartnett observed that "central banks in aggregate still printing: bought $350bn in April, $300bn in May, <$100bn in June…big 5 central banks buying less but not yet selling."
And while the ECB quietly tapered from €80 to €60BN last December (even though Mario Draghi went to great lengths to described the tapering as a non-event) and is expected to announce a formal tapering again, most likely during Jackson Hole , nowhere is this quiet slowdown in central bank purchases more evident than in the balance sheet of the BOJ, whose average purchases have declined sharply in recent months from a JPY80trillion average to a far lower level:
Indeed, it is safe to say that the topic of the liquidity injection by central banks, or rather its removal, has become one of the most discussed topics within the financial community. Case in point, in a note from last week, Credit Suisse's Andrew Garthwaite wrote that "on a year-on-year basis, the aggregate balance sheet of the big 4 central banks is set to continue expanding through 2018, although the pace of expansion will steadily slow, as shown in the second chart below. Our estimates, which include our economists' assumptions on ECB, Fed and BoJ buying and keep current FX rates constant, suggest that it is Q3 2018 when the contraction in the Fed's balance sheet on a monthly basis (which at that point will be $40bn a month) will start to exceed the purchase of assets by the ECB and BoJ."
Or, as he puts it in easily digestible format: "The inflection in central bank balance sheets comes in Q3 2018."
Today, in a comparable analysis by Citigroup, the bank's economist Kim jensen also looks at the global balance sheet and specifically the moment when it will stop growing, and also what may happen once the "unprecedented balance sheet changes" unleashed by central banks finally go into reverse, and liquidity is extracted: "We expect advanced economy (AE) net asset purchases to fall to roughly zero over the next 12-18 months, as the Fed reduces its balance sheet and the ECB, BoJ and Riksbank reduce their net asset purchases."
What happens once Q3 (or Q4) 2018 comes and the global fungible liquidity firehose of "flow" (remember it was back in 2012 that Goldman first admitted that when it comes to asset prices, only the "flow" of central bank assets matters, not the stock) is the question.
In an attempt to boost trader confidence, Citi has conducted an empirical anslysis of several "QE slowdown" episodes in the past to find that "asset markets have on balance absorbed these changes fairly well, but with significant variability across episodes." More:
Even though there has not been a “global post-QE” world before, there have been temporary slowdowns in the flows of net asset purchases in aggregate and reductions in balance sheet size for specific economies. From these episodes we may be able to glean some insight as to what we can expect as central bank stimulus through balance sheet expansion is tapered and later reversed. To that end this note considers several relevant precedents of central bank balance sheet reductions and tapers of net asset purchases, including the reductions in aggregate global central bank net asset purchases (in 2009-10, 2011, 2013/14 and since mid-2016), the reductions of the BoJ and ECB balance sheets in 2006 and 2013-14, respectively, the purchases tapering episodes for the BoE (2010, 2012) and Fed (2009, 2011, 2014) as well as the (temporary) run-down of the Fed’s MBS portfolio (2010-11) and the ECB’s Covered Bonds portfolio (in 2010-11 and again in 2012-14).
Still the are caveats...
Still the evidence suggest that financial markets did not react consistently or uniformly to the end or reversal of net asset purchases by central banks, nor can we attribute the asset market developments at the time uniquely to central bank actions; e.g. market-based measures of inflation expectations (5y5y breakevens) fell after the end of QE2 and QE3 in the US but increased after QE1 in 2009. The euro area 5y5y inflation swaps fell from 2.4% to 1.6% over the repayment window of the LTROs in 2013-14, but UK inflation breakevens rose in the three months after the end of QE1 (3.6% to 3.8%) and QE3 (3.0% to 3.5%). Indeed, in most cases and for most asset markets, the effects of CB purchases are difficult to trace and there were plenty of other relevant developments at the time that may have been more significant for asset prices than the central bank purchases.
... of which the biggest is the following: "Since 2008, there has not been a period when aggregate AE central bank balance sheets have contracted."
Well, as Credit Suisse pointed out and as Citi confirms, in roughly 12 months, the world is about to have its first period of aggregate central bank balance sheet contraction even as the flow is already shrinking at a rapid pace.
As such, the only question is when will the market finally begin to discount this contraction. For Citi the answer may be never, but others disagree: recall that to Bank of America's Michael Hartnett the answer is not 12-18 months, but 3-4 months:
We don’t think this is “big top” in stocks; greed harder to kill than fear; don’t think this “big top” in stocks, would be surprised if bull market which began with SPX 666 ends before 6666 on the Nasdaq... summer 2017 = significant inflection point in central bank liquidity trade…will likely lead to “Humpty-Dumpty” big fall in market in autumn, in our view. Big Top likely occurs when Peak Liquidity meets Peak Profits. We think that's an autumn not summer story.
And...
The most dangerous moment for markets will be when rising rates combine in three or four months’ time with an inflection point in corporate profits. In anticipation of this, we would use the next couple of months to buy volatility, and within fixed income slowly reduce exposure to IG, HY, and EM bonds.
So far Hartnett has gotten the blow off top part absolutely correct as one can easily see by what is now talking place in the market which is hitting new all time highs on a daily basis, and where downticks are no longer allowed. If he is also right about the timing of when markets will finally discount "Q3 2018", then it may be time to start quietly getting out of dodge.