---

Remember when the RBA consistently referred to the buffer of "excess savings" accumulated during the post COVID period? Have you noticed that they have largely stopped talking about it in their public communications? In this latest note, Tim Toohey, Head of Macro and Strategy, highlights that calculations of excess saving buffer are illusionary. We show that there is no longer an excess of liquid savings accumulated since COVID, and explain why the RBA should take a more pre-emptive approach to monetary policy and commence easing late-2024.

The RBA initially estimated that this excess savings buffer was $260bn, then as time progressed the estimate was revised up to over $300bn. The RBA stopped providing a running total on the estimate some time ago which is probably sensible, given the imprecision in which it is measured. Nevertheless, it was always a simple calculation, and it is worth noting that on our estimates the accumulation of excess savings recently topped a massive $500bn, or 32% of annual household disposable income.

Chart 1: Estimates of excess household saving since COVID-19 are misleading

Source: ABS, YCM.

Clearly, if households collectively decided to spend that excess savings Australia would have a massive consumption boom. Spending $500bn could sustain nominal private consumption growth at its 10-year average pace of 4.6% p.a. for four years in the absence of any income growth. To be clear, that is not a forecast. Indeed, the purpose of this paper is to illustrate that liquid "excess savings" are now exhausted and no longer available to smooth near-term consumption.

In many ways the initial observation from the chart is to ask the question as to why the stock of savings hasn't been dipped into much earlier and in much larger scale? After all, what we can say empirically is that during periods of rising household wealth the household saving ratio normally declines and given the extent of cost-of-living pressures, it is entirely reasonable to have expected households to engage in consumption smoothing.

The traditional inverse relationship between wealth and the saving ratio has indeed occurred in the past few years with the flow of household savings (i.e. the income left over post tax, interest and consumption each quarter) declining as household wealth rose. The saving rate is now proximate to zero. In theory, the household saving rate can move negative if households perceive the wealth gains to be permanent and households have few constraints to access credit. In the early 2000's this was facilitated via mortgage equity withdrawal, although other mechanisms are now possible including, as we have learnt in recent years, via early access to superannuation. That is, it is possible to sustain consuming more than you earn, for a limited period of time, at least at an individual level. But the idea that households will collectively unlock wealth gains by again using their homes as an ATM or their Super balances as piggybank to be raided and thereby drive the national saving rate collectively negative is not at present a high probability outcome. 

However, at the aggregate level, as more and more households commence the retirement phase, it may well be feasible that the household saving ratio to move negative for an extended period of time[1]. This would merely be due to retired households consuming at a faster rate than the returns being generated from their superannuation and other income generating assets. This is what the superannuation system was designed to do. The system was based under the assumption that the superannuation balance would be drawn down towards zero by the time of death. In practice, poorer households have struggled to build up meaningful superannuation balances and appear destined to be reliant on the pension system, while wealthier households with high superannuation balances have tended to consume less than the income generated from their superannuation holdings, seeing their balances continuing to compound during retirement.

For now at least, it seems we are stuck with a two-tiered superannuation system: failing to meet its objectives for the bottom half of income earners, and remaining a great wealth accumulation device for the top 20%. Whether the household saving rate moves negative as the Baby Boomers retire is difficult to know with certainty as it will be conditional on superannuation reforms, asset returns, and confidence in the system itself. However, if nothing else changes on the policy front, it is more likely that for the next few years Super will be a force that lifts the national saving rate, both by compulsion and by compounding, even if lower income households would prefer to consume their Super now rather than at some distant point in the future.

A slide from the recent CBA earnings results presentation (refer Chart 2, over page) supports this observation. Younger and more indebted households (red box) have engaged in consumption smoothing, running down their saving to fund essential spending and enabling a relatively modest decline in discretionary spending. Conversely, retirees have seen a sharp rise in their deposit savings and their discretionary spending growth has been largely unimpeded.

This CBA chart matches what we suggested would occur in our note "Cashflow Pothole in the Energy Transition Journey" (Dec 2022) where we suggested in our analysis of upcoming hits to cashflow that "retail sales will slow from the rapid rate of close to 20% (y/y) to zero growth by mid-2023", despite this being the exact period the RBA was pointing to $260bn in excess saving buffers that would sustain spending growth. It was a controversial forecast at the time, but for the record, retail sales did slow to just 1% (y/y) by mid-2023 - propped up somewhat by stronger population growth and additional government subsidies.

Chart 2: CBA cost of living analysis

Source: CBA Results Presentation, Aug 2024. 1. ABS, as at June 2024. 2. Reported by ABS as deposit and loan facilities (direct charges). 3. Including education, stamp duty and conveyancing, clothing and footwear, communication. 4. Per customer. For spending 13 weeks to end of quarter, for saving the average balance as at end of quarter. Consistently active card customers and CBA brand products only. 5. Spending based on consumer debit and credit card transactions data (excluding StepPay). 6. Includes all forms of deposit accounts (transaction, savings and term) and home loan offset and redraw balances. Trimmed mean excluding top and bottom 5% of customers within each age band. Income quartile calculated across all ages based on customers with income payments to CBA accounts in the 13 weeks to 30 June 2024, considering salary, wages and government benefits).

 

The CBA chart also accords with our note "This is Going to Hurt" (March 2023) where our analysis suggested that discretionary cashflow for the bottom 20% of households would plummet in 2023, the 35-44yo cohort had no accumulated precautionary savings post-COVID and that the vast majority of the $260bn had been accumulated by those over 45yo. As the above CBA chart shows, it has indeed been a painful period for younger and more indebted households and those with few precautionary saving buffers.

It would be a mistake to assume that younger and working age households can continue to draw upon savings to fund spending growth. Chart 3 again shows the level of excess savings since the start of COVID, but this time we exclude mandatory superannuation inflows. This is a much better measure of liquid excess savings that can be drawn upon by working households, and as of the March quarter of 2024 that buffer has now been completely depleted! If the CBA data is reflective of the broader society, then given the 65+ age group has continued to expand their savings in recent years it follows that the excess savings ex-Super since COVID for those sub-65yo must currently be negative.

Chart 3: Excess household saving accumulated since COVID-19 (excl. mandatory superannuation contributions)

Source: ABS, YCM.

The message for policy makers is that there is no longer a liquid buffer of excess household savings to draw upon, and it's likely that most households in the sub-65yo age bracket have drawn down savings buffers to below pre-COVID levels.

Deputy Governor Andrew Hauser's recent speech made it clear that the RBA are unsure of what will happen with household savings and spending in the face of recently enacted tax cuts, rising household wealth and the lagged impacts of monetary policy. That's OK, and if you believe the odds are evenly balanced then it argues for doing nothing on policy until you have a clearer picture.

For our part, we have been focussed for some time on trying to disentangle the puzzle of the impact on the saving rate from the key drivers of wage income, taxes, housing, financial and superannuation wealth and the shift in the superannuation guarantee levy. In our note "Big Super's Big Impacts" (Nov 2023) we directly modelled the savings rate and found a stable and sensible relationship between the saving rate and these key drivers. One of the key conclusions of that that study was that a rising SG levy acts as a material brake on the domestic economy's near-term growth. It concerns us that 9 months on, no one else seems aware of the dynamic yet we are rapidly lifting the SG levy whilst policy makers simultaneously consider further rate hikes.

On our modelling a shift in the Superannuation Guarantee (SG) levy by 100bps has just as large an impact upon spending as a 100ppts interest rate rise. That is, the 1 July 2024 rise in the SG levy of 0.5% to 11.5% is equivalent to a 50bp interest rate hike for the consumer. The same model suggested that over half of the 1 July income tax cut would likely be saved, particularly as it is skewed to higher income households.

We estimated that the net impact of the simultaneous rise in the SG levy and the income tax cut is broadly neutral for the aggregate consumer. Obviously under these two policies, spending that skews to older and wealthier spending categories will be a net beneficiary, whilst working age households will be mildly worse off. That is, the model suggests the RBA need not be as concerned about a post-tax cut surge in spending for most Australians. Nor should they be concerned that, should consumers suddenly turn more optimistic, there is a large buffer of liquid excess savings that can be deployed.

The RBA obviously lacks the tools to curtail the spending habits of debt-free retirees. The RBA have been clear that it wants to suppress aggregate demand, without prompting a sharp labour market deterioration, in order to return inflation to target. But the argument that the RBA needs to curtail problematic services inflation, by making room in the rest of the economy is fine until you realise that only one third of the CPI basket is in discretionary spending and just 9% of the CPI basket is discretionary services spending. Raising rates to suppress demand led inflation in just 9% of the basket is debatable in the first place, but what portion of this 9% of excess demand for discretionary services is cashed up retirees? In contrast, 37% of the CPI basket is in essential services, and rising prices for these items overwhelmingly negatively impacts lower income households. That is, excess services inflation currently has little to do with excess private demand growth of working age Australians. As the CBA chart highlights, most of the excess spending in discretionary spending is being done by those who are already in retirement and are impervious to - or even benefit from - higher interest rates. If we are looking to the RBA to curtail cashed up and retired Baby Boomers then we will end up with an unbalanced and intrinsically unfair economic system. That job clearly rests with the tax system.

It is also worth noting that per capita consumption is currently contracting at -1.1% (y/y) - the same rate as the depths of the 1991 and 1983 recessions. With population growth set to slow sharply into 2025, total consumption growth will risk resembling the current per capita growth trajectory. Together with a further SG levy increase scheduled for mid-2025 acting as an additional headwind to consumption growth, waiting until mid-2025 before considering interest rate relief could prove a costly error. In other words, we know with some certainty that key non-monetary policy decisions on immigration and superannuation will constrain future consumption and lift future savings.

For a central bank that says it no longer gives forward guidance, the irony of the RBA ruling out interest rate reductions in 2024 should be lost on no one. We agree with the Deputy Governor that policy making under uncertainty requires a healthy degree of humility and that is indeed lacking in some financial market commentators that proffer their strongly worded advice on a weekly basis. Where we disagree with the RBA current view is that they believe there is symmetry in potential outcomes for the saving rate. On our estimates, excess liquid saving buffers are already depleted, a saving rate around zero mutes the wealth effect on future consumption growth, the impact of the recent income tax cut will be neutered by the rise in the SG levy on 1 July 2024 and the certainty of the final rise in the levy on 1 July 2025 to 12% is, by definition, an event that lifts the household saving rate. In other words, there is a much greater chance that the savings rate will rise than fall in the year ahead. In conjunction with a policy-led decision to reduce population growth sharply in 2025, the implication is that the risk to economic growth is skewed to the downside.

In the Deputy Governor's own words, in situations where there is asymmetry of potential outcomes, the Bank should take a more activist or forward-looking approach to interest rates. It is precisely because of this asymmetry why we continue to suggest that the RBA should take a forward-looking activist position and commence its easing cycle in December 2024.

We conclude that there is not symmetric risk to the saving rate looking into 2025 and 2026 and this will have big implications for the outlook for economic growth, employment and inflation. The starting point is a saving rate that is already close to zero, banks are not facilitating mass access to liquidity for households to consume prior wealth gains, the superannuation system remains in a state of net inflow and the recent rapid rise in the SG levy in addition to a further rise in 2025 essentially guarantees the saving rate will rise rather than fall in the period ahead. In concert with the observations that the buffer of liquid excess savings in the post-COVID period has now been completely depleted and a clear trend rise in the unemployment rate, it is likely that a bout a precautionary saving ensues in the months ahead. 

It's clear from the RBA's current guidance that there are no rate cuts on the horizon until mid-2025. But a lot can change between now and the end of 2024. A shift in the RBA perception of the directional risk to the saving rate would be one key step in the RBA returning to be a pre-emptive central bank, opening the door to a late 2024 rate cut.