RBA gets third time lucky with dovish taper implying $147bn of QE buying
After two consecutive months of surprises provoked by the RBA trying (unsuccessfully) to shift from its prior posture of data-dependent "nowcasting" to boldly predicating decisions on unreliable forecasts of an unknowable, pandemic disrupted future, Martin Place finally did what we expected: it delivered the prudently slow, or elongated, "taper" of its bond purchase, or quantitative easing (QE) program, which is not what it might seem at surface level given we are in a short recession right now. The bottom-line is that this is a substantial increase in the stimulus the RBA is injecting compared to what it originally intended in July. On our estimates, the RBA will now undertake a minimum of circa $140bn to $150bn of bond purchases relative to the much skinnier $105bn it had in mind back in July.
Third Time Lucky
Let's back-track a little and consider how we got here. Back in February 2021 Coolabah forecast a QE3 program worth $100bn (plus more via QE4), after the RBA had delivered the first two, six month bond purchase initiatives (aka QE1 and QE2), which were each worth $100bn, or $200bn in total.
Few if any folks were expecting $100bn of QE3 in February. Yet by the time the RBA came around to making a decision on the subject---at its July board meeting---the financial market was pricing in an aggressive taper, or reduction in its bond purchase pace, such that QE3 was only going to be worth $50bn to $75bn of total investments. While there were economists forecasting $100bn or more of QE3 just prior to the July meeting (eg, Westpac's Bill Evans), market pricing was much more hawkish.
Our argument in July was that maintaining the existing weekly bond purchase pace at $5bn was the prudent path given:
- the major central banks, and the Fed and ECB in particular, had not announced their own tapers, and the RBA had consistently guided the market to believe it would be the last of the central banks to tighten policy;
- the RBA had conditioned the market to believe that monetary policy was being predicated on nowcasting rather than its historically rubbery forecasting, and while the economic data had surprised on the upside, Delta had forced NSW into lockdown, and the RBA was still miles below its wages and inflation targets; and
- key media proxies were signalling that the RBA would maintain a $5bn/week purchase pace and then taper down to $4bn/week in November.
Embarking on any taper, and certainly an aggressive taper, seemed imprudent. Yet at its July meeting the RBA did indeed announce that in September it would start reducing its purchases down from $5bn/week to $4bn/week, and then review them at its quarterly forecasting meetings in November, February, May, and so on.
The RBA confused economists and market participants by describing the taper as very much a "line-ball" call at its July meeting, and repeatedly advising that it would increase the QE program if there was any adverse news on the economy vis-a-vis its super-optimistic central case. (In the RBA's defence, they have done exactly that---increase QE3 in response to a shock---by extending it rather than via increasing the weekly purchase pace.)
And then Delta duly forced the nation into recession, the dovishness of which was only amplified by the fact that the pre-lockdown data on both wages and inflation materially disappointed to the downside. That is to say, even though the Aussie economy was running hot heading into lockdown, there was no evidence of wage or inflation pressures emerging, as many had hoped.
What the RBA has not publicly acknowledged is that a key motive for embarking on the pre-emptive taper of its bond purchases was an arguably unjustified concern that maintaining the $5bn/week run-rate of bond purchases into early next year would have resulted in the RBA owning 40-50% of the individual bonds it was buying, which would have impaired market liquidity.
Some central banks regard this as the upper bound on QE, which the RBA was evidently keen to avoid bursting through, especially if it was forced to prolong QE---as it has been---by negative economic shocks.
It turns out, however, that new data released by APRA this week shows, as Signal Macro's Matt Johnson has highlighted, that while the RBA was buying government bonds, banks have been selling them in massive volumes, and overall market liquidity has increased, not decreased as many banks have incorrectly claimed (we have publicly and privately argued that liquidity has been excellent). Quoting Matt:
That’s exactly what we see in the recent APRA and AOFM data. Bank holdings of government bonds are down and market liquidity is up. APRA data shows that...the share of the Commonwealth bonds market held by larger (LCR) banks has fallen from a peak of 23%, in Q3'20 to around 10%. The larger bank share of the State government bond market has fallen from 49% to 44%.
Of course, some of this is because larger banks sold bonds to the RBA in QE. However, even after adjusting the outstanding stock of bonds for the RBA holdings, their holdings have declined by 100bn (or 6ppts to 28%, note this is the combined Commonwealth bonds + State government bond market, adjusted for RBA QE purchases).
While it's true that larger banks held more than 30% of the (QE adjusted) market in 2020, it didn't reduce market liquidity. Turnover in Commonwealth bonds peaked at a an annualized pace of ~350% in Q1'2020 (turnover spikes are common in shocks), and remained around 300% all year. So by this measure, bonds were more liquid in 2020, when banks held a larger share of the market. The trend is probably the most important thing for regulators: this shows that turnover has risen from around 260% prior to 2020 to around 300% today.
Confusing Eradication with Endemics
The RBA also surprised market participants by releasing heroically hopeful forecasts in August that actually upgraded its May and July estimates of the economy's future path, and attributed zero long-term impact to the current lockdowns that have shuttered about 60% of the country.
It was hard to understand why the RBA would ignore the data we had on the effects of Delta locally and globally, and assume that the escape from lockdowns into the zero COVID world in 2020---accelerated by the tail-wind of much larger fiscal and monetary stimulus programs---would be emulated in our transition out of the 2021 lockdowns into an "endemic" world where we have to slowly adapt to living with COVID (and substantially skinnier stimulus).
It is likely that there has been permanent scarring of both individuals and businesses from the experience of repeated lockdowns (pity the poor folks living in the most locked-down city in the world, Melbourne). And we will be collectively more risk-averse as we exit lockdown in 2021 and slowly step into 2022. Having said that, we will bounce-back, albeit to a new normal characterised by more cautious behaviours.
So while we do have cause to be constructive about the future, the RBA's task of getting the economy to full employment characterised by wage growth that can push inflation back into its target band has arguably only got harder.
The Taper is Delivering Much More Stimulus
The confusion around what the RBA would or would not do with QE was exacerbated by the fact that the capacity constraints in the current program are all of the RBA's own making, and as others have noted, easy to fix. The RBA could have extended the maturities it was buying and recalibrated the the mix.
But in its characteristically rigid fashion, it has chosen not to do so, and would prefer to resolve these problems through other, more indirect means, which brings us back to what it decided regarding QE3 this week.
What we have now is a much bigger QE3 purchase program, which is entirely appropriate for the circumstances. Notwithstanding my comments above, QE has been hugely successful, smashing down both 10 year Aussie interest rates and the Aussie dollar, much more effectively than anyone, including the RBA, ever imagined.
And in contrast to reductions in the RBA's overnight cash rate, which contribute to lower variable-mortgage rates, buying 5-year to 10-year government bonds does not directly impact the housing market much, and therefore risk blowing housing bubbles, as the RBA has historically done when it slashed its overnight cash rate (cf., the booms/bubbles in 2003 and 2004, between 2012 and 2017, and commencing again in 2019 to the present day).
QE is much safer from a financial stability perspective, furnishing the RBA with a powerful new tool that it can use to deliver a more balanced monetary policy platform that adjusts both short- and long-term interest rates.
The final QE3 program that the RBA revealed this week involves it dropping the quarterly bond purchase pace from $5bn/week to $4bn/week, but crucially maintains that all the way through to February rather than further reducing it to $3bn/week at its next quarterly meeting in November. This emulates the Bank of Canada's approach, which did exactly the same thing when it first tapered QE.
Central banks have discovered time and time again that there is zero upside from fast tapers and lots of downside.
We have previously referred to this new "slow taper" from the RBA, which we expected, as the 4/4/3/2/1 schedule where that denotes the RBA buying at $4bn/week for the first two quarters and then trimming by $1bn/week each quarter thereafter. (This also follows the Bank of Canada.)
This generates total bond buying of $147bn (see our note on this here), which is much bigger than the market's original estimates for QE3 in June/July (ie, between $50bn and $75bn) and the mid-point of economist estimates, which ranged from $50bn to $150bn at the time (back in July CBA had $50bn pencilled in for QE3, UBS estimated $75bn, and Westpac's Bill Evans fingered $150bn).
Interestingly, it is almost identical to what Coolabah expected back in July, which was a 5/4/3/2/1 program that resulted in a total of $157bn of buying (or $149bn if the RBA went with a more aggressive 5/4/3/2/0).
So the RBA is providing a lot more stimulus than it originally intended, and sensibly responding to the adverse news on Delta. And it is increasing, rather than reducing, the support it is providing to the economy. (For what it is worth, amongst the journo's, I would say that Sophia Rodrigues, James Glynn, and Terry McCrann (more subtly) got this right.)
It was, for example, possible that back in July the RBA had in mind a QE3 taper than looked like a 4/3/2/1 or even 4/3/2/0 schedule, which would have resulted in about two-thirds of the bond buying it will now do in QE3 (specifically $105bn and $92bn respectively).
If the RBA were to get ahead of itself again, and run with a more accelerated 4/4/3/2/0 schedule (ie, drop off the last quarter of $1bn/week of buying), it would still end-up purchasing $139bn, which is appropriately more stimulatory than its original July plan (ie, circa $100bn on a 4/3/2/1 schedule). The problem with this more truncated path is it reduces the RBA's optionality if something goes wrong in Q3 or Q4 next year...
More Shocks are Possible
A few final points to consider. First, more negative economic shocks are entirely possible over the next 12 months, which could convince the RBA to further extend QE3. And contrary to popular myth, the RBA has loads of ammo up its sleeve. It can extend the bond maturities it is buying, and toggle the mix between Commonwealth vs State government bonds.
It could also reintroduce the enormous bank subsidies delivered via its $188 billion Term Funding Facility, whereby it lent at a cost of just 0.1% annually to the banks for a term of 3 years.
Now that is a licence to print money and one reason why the banks are suddenly so wildly profitable again. And given the $1.5 trillion in government bonds and cash on deposit at the RBA, it reinforces the question Signal Macro's Matt Johnson and I keep on asking regarding why APRA and the RBA are continuing to allow banks to hold $140bn of their own loans and one another's bonds instead of holding government bonds in what is (by the banks' own admission) a huge subsidy from taxpayers to the banking system (banks get a lower cost of funding and higher returns on equity while taxpayers have to pay higher interest rates on their bonds). As Johnson writes here at Livewire today:
Recent data suggests that APRA and the RBA can shrink the Committed Liquidity Facility (CLF) faster than I had previously figured. The entire point of the CLF was to protect the liquidity of the government bond market. Given that banks have been voluntarily selling government bonds, it follows that the CLF is too big. The ongoing increase in turnover also supports the case for higher bank ownership of government bonds, and hence a faster closure of the CLF. I now think they can half it in 2022, and close it in 2023.
The second problem for the RBA is that there is no evidence it is remotely close to fulfilling its full employment and inflation targets, judging from the wages and core inflation data. With the cash rate at its 0.1% lower bound, QE offers it probably the most policy optionality right now in terms of continuing to inject stimulus without blowing bigger housing bubbles if it discovers in 2022 that it is continuing to fail to meet its legislative objectives, which it has struggled with for a long time now. It is therefore easy to imagine scenarios whereby QE3 is much larger than currently planned, although we should all hope that they do not come to pass...
Finally, some say the RBA does not like QE because it looks like it is funding the government. In 2021 this is a pretty silly suggestion. Aussie government bonds are amongst the most sought after globally. Aussie State government bonds pay the highest interest rates of any AAA or AA rated government bonds in the world hedged into AUD, USD, EUR or JPY. (That's almost certainly because Aussie banks are buying their own bonds and loans under the CLF instead of holding government bonds for emergency liquidity as is best practice overseas.) So Aussie governments can finance as much debt as they want today.
The RBA is only focused on influencing the interest rates they pay on this debt, which impacts the costs taxpayers bear, and which serve as the benchmark interest rates for all borrowing across the economy. QE is much more effective at influencing those rates than just using the very limited overnight cash rate. And because of its success, it has become a conventional part of the RBA's toolkit.
Access Coolabah's intellectual edge
With the biggest team in investment-grade Australian fixed-income and over $7 billion in FUM, Coolabah Capital Investments publishes unique insights and research on markets and macroeconomics from around the world overlaid leveraging its 14 analysts and 5 portfolio managers. Click the ‘CONTACT’ button below to get in touch.
Chris co-founded Coolabah in 2011, which today runs $7 billion with a team of 33 executives focussed on generating credit alpha from mispricings across fixed-income markets. In 2019, Chris was selected as one of FE fundinfo’s Top 10 “Alpha...