Speech
A Review of the RBA’s Term Funding Facility
Christopher Kent
[
*
]
Assistant Governor (Financial Markets)
Sydney
–
9 October 2024
Audio
34MB
Watch video: A Review of the RBA’s Term Funding Facility
Thank you for coming to the Reserve Bank’s offices today. I will talk about a review we have
published on the Term Funding Facility (TFF). 1
This is the fourth instalment of the series of reviews
of unconventional policy tools the RBA used during the
COVID-19
pandemic. 2
In March 2020, the economic outlook was bleak and highly uncertain (Graph 1), financial markets were
in turmoil, and there was limited scope to lower the cash rate further. In that environment, the RBA
pursued a package of policies to support the economy. 3
The TFF review considers how that element of
the package worked, whether it achieved its aims, and lessons for the future. I will cover the key points
but there is a lot of detail in the review itself.
Graph 1
What was the TFF intended to do?
The TFF aimed to:
lower the cost of borrowing for businesses and households, by lowering lenders’ funding costs,
and to reinforce the benefits to the economy of the lower cash rate
encourage banks to lend to businesses – particularly small and medium-sized enterprises (SMEs)
– given that business credit tends to fall in downturns.
How did it work?
The TFF provided low-cost three-year funding to banks, which also indirectly helped to lower the cost of
borrowing from wholesale markets.
Under the TFF, banks had access to cheap funding up to an amount that was based on the initial size and
subsequent growth of their loan book. The interest rate was initially fixed at 0.25 per cent.
This was lowered to 0.1 per cent in step with the reduction in the cash rate target in November
2020. A bank was able to secure additional TFF allowances if it increased its overall lending to
businesses, particularly smaller businesses. For each dollar of additional credit extended to large
businesses, a bank was eligible for another dollar of TFF funding. For each additional dollar extended to
SMEs, a bank had five more dollars added to its TFF allowance.
Banks could access their allowances up to the end of September 2020. However, by the time of the September
Board meeting, the economy was still far from the RBA’s goals, and considerable downside risks
remained. The Board decided to extend the facility and increase banks’ allowances; banks could
access their new allowances for three-year fixed-rate loans until mid-2021. 4
TFF funding was much cheaper than other sources of term funding. Unsurprisingly, banks took up most of
their TFF allowances (Table 1). 5
The TFF ultimately provided $188 billion of
funding, which was equivalent to 6 per cent of the stock of credit outstanding at the peak of
the TFF’s use. Banks repaid all TFF funds as scheduled by mid-2024 without incident.
Table 1: TFF Usage Across Banks
Amount drawn
$ billion
Share of total allowances
Per cent
Major banks 133
100.0
Mid-sized banks 24
99.6
Small banks 9
58.3
Foreign banks 22
54.2
Total across all banks 188
88.3
Sources: APRA; RBA.
To summarise its effect on funding costs for banks and others with access to wholesale funding markets:
The TFF lowered banks’ funding costs directly. For the major banks, the TFF was around
60 basis points cheaper than issuing bonds during the TFF drawdown phase (Graph 2). It
lowered their average cost of funds by around 5 basis points. 6
Graph 2
Together with other parts of the policy package, the TFF also indirectly helped to lower the cost of
wholesale funding. With the TFF in place, banks had little need to issue bonds but investor demand
for those and other similar securities remained strong. Strong demand coupled with a sharp fall in
supply contributed to a decline in yields on a range of existing and newly issued securities. This
included securities issued by non-major banks (which continued to issue bonds). Non-bank lenders also
benefited significantly; their issuance of residential mortgage-backed securities (RMBS) – a
key source of their funding – picked up significantly as the cost of issuance dropped sharply
(Graph 3). 7
Graph 3
Did the TFF achieve its aims?
Banks passed lower funding costs through to retail lending rates for both households and businesses, on
both new and outstanding loans. On average, outstanding lending rates fell by almost 100 basis
points – a little more than the 84 basis point decline in banks’ overall cost of funding
(Table 2). 8
The fall in business rates was comparable across
variable- and fixed-rate loans, with larger reductions for SMEs than was the case for larger businesses.
But the fall in mortgage rates was much more pronounced for fixed-rate loans; the decline in fixed rates
was also large relative to the reduction in the cash rate compared with earlier episodes of monetary
policy easing. Banks’ decisions to provide fixed-rate mortgages at very attractive rates was
consistent with the low fixed-rate TFF loans as well as banks choosing to focus their competitive efforts
in the fixed-rate mortgage market.
Table 2: Changes in Funding Costs and Outstanding Lending Rates
February 2020 – February 2022
Change
Basis points
Cash rate target
−65
Funding costs
(a)
−84
Overall mortgage rates
−97
– Variable mortgage rates
−68
– Fixed mortgage rates
−152
Overall business lending rates
−105
– Variable business lending rates
−103
– Fixed business lending rates
−89
(a) Major banks.
Sources: APRA; ASX; Bloomberg; LSEG; major bank liaison; RBA.
Households and businesses that took out fixed-rate loans benefited from the particularly low fixed rates
on offer at the time. The share of new housing lending at fixed rates rose from around
15 per cent at the start of the pandemic to a historical high of over 45 per cent by
mid-2021. 9
Not only were existing borrowers switching from
variable to fixed rates, but new mortgage lending also picked up noticeably through 2020 and into 2021
(Graph 4). In addition, lower rates contributed to a pick-up in disposable incomes of debtors. In
these ways the TFF (together with other parts of the policy package) helped to support dwelling
investment, the housing market more broadly, and other elements of aggregate demand.
Graph 4
The TFF was also intended to support the availability of credit. We were particularly concerned that banks
might have been reluctant to continue to extend credit to businesses during such difficult times. The TFF
is likely to have played a role in underpinning business credit. It encouraged demand by contributing to
lower rates for borrowers. It also encouraged banks to expand lending to businesses to obtain additional
low-cost TFF loans. Indeed, business credit held up better during the pandemic than in the global
financial crisis (GFC) (Graph 5); such declines had also been evident in earlier downturns. Despite
the supporting role of the TFF, total business credit may not have increased through 2020 and 2021 for
several reasons, including a lack of business confidence and the reduced need for business credit given
the sizeable government support to businesses’ cashflows. And despite the considerable incentives in
the TFF to expand SME lending, staff estimates found no statistically significant effect on total SME
lending compared with large businesses.
Graph 5
While not an explicit goal, one other benefit of the TFF was the indirect support it provided to the
public sector balance sheet. By supporting stronger economic outcomes, the TFF – together with
other monetary policy measures – contributed to higher tax revenues and lower support payments to
households and businesses than would otherwise have been the case.
How much did the TFF cost?
The TFF was part of the insurance the RBA took out against a catastrophic economic outcome. While some of
the TFF’s design features underpinned its significant use by the banks – and hence its
economic benefits more broadly – these were also associated with financial costs for the RBA. The
total cost to the RBA is estimated to have been $9 billion. There were several reasons for this
cost.
First, the choice to supply funds at a fixed rate was intended to give banks and their borrowers
certainty, thereby reinforcing the other elements of the policy package: notably the RBA’s
three-year yield target, and its forward guidance. But the economic recovery and increase in inflation
turned out to be much stronger, and started much earlier, than the initial upside scenarios considered by
most economists and the RBA. As a result, the Board ended up raising the cash rate target by much more
and much sooner than had been expected (Graph 6). While the TFF was profitable for the RBA until May
2022, once the cash rate increased, the RBA was paying banks more interest for the balances that they
kept at the RBA than the low fixed rate the banks were paying on the TFF. Because the banks passed these
lower funding costs in full, household and business borrowers who had locked in low fixed rates were the
ultimate beneficiaries as interest rates rose.
Graph 6
Second, around $4 billion of this cost was the result of the Board’s decision to extend the TFF
in early September 2020. At that time, the banks had taken up just 60 per cent of their initial
TFF allowances, with almost half of that occurring as late as August (Graph 7). This suggested that
the banks did not need TFF funding to compete for, or satisfy, the demand for borrowing from households
and businesses. Rather, the banks waited until as late as practical to draw down TFF funds because doing
so extended the time the TFF would contribute to meeting regulatory liquidity requirements on the
banks. 10
A similar pattern of late take-up was later observed
with the second tranche of the TFF.
Graph 7
Some lessons for the future
The TFF delivered on its goals. It lowered borrowing costs for a range of borrowers, kept credit flowing
to the economy, and supported aggregate demand. In addition, along with other measures – including
the purchase of bonds in the early weeks of the pandemic – it helped to restore confidence in
financial markets, which were significantly disrupted in the early days of the pandemic.
Based on the findings of the review, the Board judged that a term lending tool of this kind would be worth
considering again if warranted by extreme circumstances. But there were valuable lessons we learnt along
the way that could help to shape any future program of this type.
Degree of support for the economy versus flexibility
Central banks can choose between fixed- or variable-rate facilities. The fixed-rate option was chosen for
the TFF in part to reinforce other policies: the yield target and forward guidance. Such policy packages
can be particularly valuable when the standard interest rate lever is already near zero and significant
downside risks to the economy remain. But the flipside to a fixed-rate facility is that it lacked
flexibility. And given the large take up of the TFF at a very low fixed rate, it incurred a material
financial cost to the RBA when the economic recovery and pick-up in inflation turned out to be much
stronger, and started much earlier, than had been expected.
Indirect effects
Many non-bank lenders were concerned that the TFF would undermine their competitive position
vis-à-vis the banks. We had expected the TFF to help lower rates in wholesale funding markets to a
degree. 11
But this effect was much stronger and more
pervasive than we had anticipated. The TFF helped to lower funding costs significantly for a range of
lenders and corporations that had no access to TFF funds. 12
It is hard to identify the
specific contribution of the TFF to these lower funding costs separately from the effects of the
wider policy package. But staff estimates suggest that these indirect effects caused yields on RMBS
to be around 50 basis points lower than they would otherwise have been.
Open lines of communication between the RBA, other government agencies and industry
Another lesson is the importance of collaboration with other government agencies, and regular contact with
industry participants. Collectively, this helped financial stability risks associated with the TFF to be
well managed. This included monitoring and managing banks’ refinancing and liquidity needs well
ahead of the repayment of their TFF loans, although that task could have been more challenging under less
favourable market conditions.
Similarly, for household and business borrowers, the RBA, the Australian Prudential Regulation Authority
and the banks’ close monitoring (and banks’ prudent lending standards) helped to reduce the
risks associated with the rise in borrowers’ mortgage payments when their very low fixed rates
rolled over to much higher variable rates. Only a very small share of borrowers struggled to meet the
increase in their mortgage obligations when their low fixed rates expired.
Importance of contingency planning, risk mangement and governance
One of the important lessons is the value of planning ahead and being ready for a wide range of
operational contingencies. We got the TFF up and running quickly in part by relying on existing,
well-understood practices. But the speed with which the RBA designed and implemented the TFF also limited
our ability to fully consider and manage the associated risks.
Forward planning can expand the options available, help us to better weigh up the costs and benefits
of each, and prioritise any pre-emptive operational work. On this latter point, for example,
floating-rate term-lending would have been challenging for both the RBA and the commercial banks to
adopt in early 2020, because neither the RBA nor the banks were readily able to undertake
floating-rate repos. The RBA and the banks have since upgraded systems and now have the capacity to
easily undertake either floating- or fixed-rate repos.
Design features could have competition implications. While RBA staff liaised with the Australian
Competition and Consumer Commission during the TFF’s setup, it would be helpful to consider
competitive implications ahead of time for any future facilities.
Finally, and perhaps most importantly, the Board has agreed to strengthen the way it considers risks,
including by examining a wide range of economic scenarios when making policy decisions involving
unconventional tools, and how to judge appropriate exit paths from such tools. In retrospect, a
greater focus on potential upside economic outcomes could have led to a different calibration of the
TFF, including deciding not to extend it in September 2020.
Summing up
The TFF met the objectives we set out for it at the start of the pandemic. It helped prevent dire economic
outcomes at a time when the outlook was bleak and highly uncertain, and there was limited scope for
further cuts to the cash rate. The TFF contributed to materially lower lending rates for households and
businesses by reducing funding costs directly for banks, and indirectly for other institutions that
borrow from wholesale funding markets. It kept credit flowing to households and businesses at a time when
banks might have otherwise curtailed lending. In helping to prevent a much more severe economic downturn,
the TFF also contributed to stronger public sector balance sheets than otherwise.
Would the RBA use a term-lending tool again in the future? The Board would consider such a tool in extreme
circumstances when the cash rate target had been lowered to the full extent possible. But it would do so
only after consideration of a wide range of scenarios and the associated risks, and with a broader range
of operational options than were available at the time of the pandemic.
What’s next?
In line with recommendations from the Review of the RBA, we will be publishing a framework for additional
monetary policy tools next year. 13
The broader set of lessons learned from the combined
use of a range of unconventional monetary policies will be considered as part of that framework.
Endnotes
I thank Susan Black, Iris Chan, Tekla Bastian,
Venura De Zoysa, Siobhan Hutchings, Ben Jackman, David Meredith, Max Prakoso and Peter Wallis for
their assistance in helping me to prepare this speech.
[*]
The full review is available at RBA (2024),
‘
Review of the Term Funding Facility
’, October. 1
The series of reviews are available at RBA,
‘
Reviews of the Monetary Policies
Adopted in Response to
COVID-19
’. 2
The package comprised a lower cash rate, a
commitment by the Board to not raise the cash rate target until it was confident that inflation
would be sustainably within the target range, the introduction of the TFF, bond purchases to
address market dislocation, and a yield target on three-year Australian Government bonds: see
Lowe P (2020), ‘
Statement by Philip Lowe,
Governor: Monetary Policy Decision
’, Media Release No 2020-08, 19 March. 3
See Lowe P (2020), ‘
Statement by Philip Lowe, Governor: Monetary Policy
Decision
’, Media Release No 2020-20, 1 September; RBA, ‘
Term
Funding Facility Increase and Extension to Further Support the Australian Economy
’,
1 September. 4
TFF usage was high by comparison with similar
schemes internationally. 5
The cost differential to obtain new funding
through the TFF would have been even larger for non-major banks, which typically have higher
issuance cost because of their lower credit ratings. 6
Securitisation markets were also supported by the
Australian Office of Financial Management’s (AOFM) Structured Finance Support Fund (SFSF),
which provided support to non-banks and smaller banks through public and private investments in
structured finance market: see AOFM, ‘Background and Legislation’, available at
<https://www.aofm.gov.au/securitisation-investments/background-legislation>. 7
Indeed, banks’ net interest margins declined
over this period due to a range of factors, including strong competition and lower
spreads-to-swap on fixed-rate loans. 8
Refinancing by households from variable-rate
loans to lower fixed-rate mortgages was also elevated. 9
See RBA, n 1, for a detailed discussion of the
TFF and banks’ liquidity requirements. 10
See, for example, Kent C (2021), ‘
The Term Funding Facility, Other Policy Measures,
and Financial Conditions
’, Address to KangaNews, Online, 9 June. 11
In particular, non-banks that specialised in
non-prime lending (and hence did not typically compete with the banks) benefited. However, some
non-banks – particularly those that compete directly with banks in housing lending markets
for prime borrowers – noted that non-banks did not receive the same level of policy support
through the SFSF as banks using the TFF, which was significantly cheaper than issuing in
securitisation markets. 12
See Australian Government (2023), ‘
Review
of the Reserve Bank of Australia
’, Final Report, March.