Has the reaction function changed?

Research evidence excerpt

Has the reaction function changed?

IdeaMperiod of economic slack to stabilise inflation.

How does this square up with our MS SAMM?

We have tested the reasonableness of these QPM dynamics in our own macro

model. We set the nominal neutral rate to 5.5% (as per the SARB), which implies

that the current policy rate already represents approximately 150bp of tightness. In

response to a 150bp deterioration in inflation expectations (from 3.0%Y to 4.5%Y),

our model requires cumulative rate hikes of approximately 175-200bp to stabilise

inflation at target by the end of 2028. This implies a repo rate of around 7.25-7.50%,

or a further 25-50bp of tightening from current levels. For what it's worth, our

actual baseline for inflation expectations is 3.8%Y in 2026 and 3.6%Y in 2027.

Both models therefore require policy to remain materially tighter than it otherwise

would. The difference is that MS SAMM generates a somewhat higher and earlier

peak. This reflects its lower degree of interest-rate smoothing and, importantly, the

greater weight we place on realised CPI in driving expectations.

Where to from here?

We do not interpret the July hold as evidence that the SARB’s reaction function

has changed. As we proposed in our post-meeting review, it would appear as if the

committee prefers to run tighter policy for longer as the growth and inflation

implications of the supply shock play out.

The seemingly benign response to a material deterioration in inflation expectations

needs to be interpreted cautiously. Relative to the baseline, the scenario removes

anticipated easing, raises the policy rate by approximately 50bp and maintains a

restrictive stance for substantially longer. The duration of the response accounts for

an important part of the model’s disinflationary effect.