Recession Heuristics

The following summarises simple recession indicators for Australia. Data is included through to .

Sahm Rule

Or:

where is the 3 month moving average of the seasonally adjusted national unemployment rate.

There are a number of permutations to the rule using different labour market measures:

• Original Sahm rule: uses the unemployment rate;
• Alternative Sahm rule: uses the employment to population ratio, because recent labour market changes such as access to benefits and work tests have reduced the relevance of unemployment as a measure of economic health;
• Vacancy rate test looks at the change in job openings, for Australia the ABS publishes this data quarterly;
• Hours worked as a proportion of potential hours for full and part time employees, combines reductions in hours as employers reduce demand, the measure has additional smoothing;
• Underemployment rate, the proportion of the labour force that is unemployed, plus those looking for additional hours;
• Underutilisation rate, the proportion of the labour force that is underemployed plus persons actively looking for work but not available to start immediately;
• Michez test uses the minimum of the unemployment and vacancy rate series.

The vacancy rate and employment to population ratio rules are calculated inverse to the unemployment rate:

where is the vacancy rate, the number of job vacancies divided by all employed persons (plus vacancies).

Although the original Sahm rule used the trailing 12 months (excluding the current month), consistent with the Michez rule the calculation here uses the last 13 months (including the current month). This index can never be negative.

Output Gap

The simplest measure of the output gap is the difference between long term trend GDP and the latest GDP. Typically trend GDP is calculated using theHoderick Prescott (HP) Filter which has allows adaptions in the trend for changes in long term growth rates.

where is the Hoderick Prescott trend filter and is a penalty term which governs the smoothness of the filter (for quarterly data typically set to 1600). The implementation below uses the one-sided filter which ensures trend is based only on the information currently available (not future observations).

The output gap was strongly negative during the pandemic period, as the economy opened up after 2022 GDP was growing faster than trend. Since the beginning of 2024 the gap is increasingly negative consistent with the deterioration in the Sahm rule measures based on labour market conditions.

Taylor Rule

The Taylor Rule is a simple policy rule which combines output (GDP) and inflation and provides a rule of thumb guidance on interest rates settings for monetary policy purposes. It is calculated as:

where:

• is the current level of annual inflation (as measured using the GDP deflator) and is the deviation from the target inflation rate (for example 2.5% being the mid of the 2 to 3% target band);
• is the current level of output (in chain volume terms) and is the Output Gap;
• is the natural level of long run interest rates;
• and are weights applied to deviations in output and inflation respectively, the original rule suggested values of 0.5 each.

Different monetary policy regimes can be more or less focused on infaltion deviations which can be simulated through selecting values for and using the sliders:

The Taylor Rule currently suggest interest rates are close to their required level from a long run perspective, although this largely depends on the assumed rate of required long run real interest rates.

Term Spread

Term spreads, the difference between short and long term interest rates, indicate the market's expectations for future policy rate changes. The term spread is calculated as the difference in the nominal interest rates. A number of different horizons can be used, two common definitions are:

• the long horizon term spread: the difference between the 10 year fixed and 1 year fixed :
• the short horizon forward spread: the difference between the 3 month bill rate 18 months forward and the current 3 month bill rate.

The implied recession probability is derived from a probit regression by the US Federal Reserve and applied crudely to Australian term spreads as:

Using the US Federal Reserve estimates a flat yield curve (the 10 year government bond rate equals the overnight rate) is associated with a recession probability of 30%. The more inverted the yield curve the greater the probability of recession from this simple model.

The recession probability produced from the yield spread, as it reflects the occurrence of recessions in the US, has a significantly higher probability than estimates calibrated to Australia (at least as measured using data for the last 30 years).

Confidence

Consumer and business confidence surveys provide an insight over time into consumer and business sentiment which has been shown to be a relatively strong indicator of future economic growth. The following charts provide the latest confidence estimates from the OECD. The OECD also publishes a composite leading indicators index which combines a range of indicators across sentiment and real economic activity. All indicators are detrended and standardised, when less than 100 the indicators are negative and imply slower future economic growth.