Forward-looking adjustments to a provision matrix

Historical loss rates describe the economy of the past few years. If the outlook is worse, or better, the provision matrix must change. This guide explains how companies link loss rates to economic indicators, how to calculate a forward-looking scalar, what to do when no relationship can be found, and works through an example.

By Hamza Fida, Chartered Accountant. Reviewed by Mirza Fahad Baig, Chartered Accountant. 3 minute read.

Short answer

A forward-looking adjustment updates historical loss rates in a provision matrix for current conditions and forecasts, as IFRS 9 requires. Companies usually find an economic indicator that has moved with their loss rates in the past, such as GDP growth, unemployment or an industry measure, and scale the historical rates by how far the forecast differs from the historical average. In this guide's example, GDP growth forecast at 0% against a historical average of 2% raises loss rates by 15% and the allowance from 25,180 to 28,957.

At a glance

Required by
IFRS 9: current conditions and forecasts
Typical indicators
GDP, unemployment, industry data
Method
Scale historical loss rates
Evidence
Past relationship with loss rates
No relationship found
Judgement, still evidenced
Excel
Loss rate builder
Forward-looking adjustments to a provision matrixRequired by: IFRS 9: current conditions and forecasts; Typical indicators: GDP, unemployment, industry data; Method: Scale historical loss rates; Evidence: Past relationship with loss rates; No relationship found: Judgement, still evidenced; Excel: Loss rate builder.KEY FACTS AT A GLANCEForward-looking adjustments to a provision matrixRequired byIFRS 9: currentconditions and forecastsTypical indicatorsGDP, unemployment,industry dataMethodScale historical lossratesEvidencePast relationship withloss ratesNo relationship foundJudgement, stillevidencedExcelLoss rate builderTax BakersForward-looking adjustments to a provision matrixRequired by: IFRS 9: current conditions and forecasts; Typical indicators: GDP, unemployment, industry data; Method: Scale historical loss rates; Evidence: Past relationship with loss rates; No relationship found: Judgement, still evidenced; Excel: Loss rate builder.KEY FACTS AT A GLANCEForward-looking adjustments to aprovision matrixRequired byIFRS 9: current conditions and forecastsTypical indicatorsGDP, unemployment, industry dataMethodScale historical loss ratesEvidencePast relationship with loss ratesNo relationship foundJudgement, still evidencedExcelLoss rate builderTax Bakers
Key facts at a glance, as set out in this guide.

Why does a provision matrix need a forward-looking adjustment?

IFRS 9 measures expected credit losses using reasonable and supportable information about past events, current conditions and forecasts of future economic conditions. Historical loss rates cover only the first. A company entering a downturn with loss rates from three good years would understate its allowance unless it adjusts them.

  1. Choose candidate macroeconomic or industry indicators relevant to the customers: GDP growth, unemployment, consumer spending, an industry output index, or interest rates.
  2. Compare their history with the company's loss rates over the same periods, ideally with a lag, because defaults follow downturns.
  3. Estimate the relationship: for example, loss rates rise by 7.5% for each percentage point fall in GDP growth.
  4. Apply it to the forecast for the next 12 months or the life of the receivables.

A forward-looking adjustment, worked through

Average GDP growth during the historical period was 2.0%; the forecast for next year is 0.0%. The scalar is 1 + 7.5% x 2.0 = 1.15.

BucketBalance todayHistorical loss rateAdjusted loss rateAllowance
Current900,0000.50%0.57%5,164
1-30 days180,0002.49%2.87%5,162
31-60 days60,0006.99%8.04%4,825
61-90 days25,00017.98%20.68%5,170
Over 90 days15,00050.06%57.57%8,636
Total1,180,0002.45%28,957
Allowance by bucket: historical vs forward-lookingAllowance by bucket: historical vs forward-looking4,4905,164Current4,4895,1621-30 days4,1964,82531-60 days4,4965,17061-90 days7,5098,636Over 90 daysHistorical ratesAdjusted rates
A 15% uplift across the matrix.

In this example the adjustment increases every bucket by the same proportion. The adjustment adds 3,777 to the allowance. The Forward-looking sheet of the Loss rate builder (Excel) recalculates it for any forecast.

What if no relationship with the economy can be shown?

Small companies often lack enough history to prove a statistical link between their losses and the economy. IFRS 9 does not require an elaborate model, only reasonable and supportable information available without undue cost or effort. A company can use qualitative evidence, such as customers' payment behaviour in the most recent months, industry reports or known problems at large customers, and make a reasoned adjustment, documented with that evidence. Concluding that no adjustment is needed is also acceptable in some cases, if the outlook is genuinely similar to the historical period and the reasoning is recorded.

How often should the adjustment be updated?

At every reporting date, using the latest forecast available. Half-year and quarterly reports need the adjustment refreshed too, because the outlook can change quickly; carrying forward last year's scalar without review is a common audit finding.

Should every bucket be adjusted by the same scalar?

Not necessarily, although most companies do. A downturn may hurt current balances proportionally more, as more good customers start paying late, while balances already over 90 days past due are mostly lost either way. Companies with enough data sometimes use different sensitivities by bucket. A single scalar is simpler, easier to audit and the most common approach.

Common mistakes

  • Leaving the matrix unadjusted while the company's own forecasts, used elsewhere, show a weakening economy.
  • Adjusting for the economy and also for recent arrears trends that already reflect it, double counting.
  • Using national GDP when the customers are concentrated in one sector or region that behaves quite differently from the economy as a whole.

See historical loss rates, forward-looking information in ECL and the IFRS 9 provision matrix.

Need help applying the standards?

Our Chartered Accountants help finance teams and students apply IFRS and US GAAP to real transactions.

Book a call Chat on WhatsApp

Questions people ask

What is a forward-looking adjustment to a provision matrix?

An adjustment of historical loss rates for current conditions and forecasts, usually by scaling them for an economic indicator such as GDP or unemployment.

How do you calculate a forward-looking scalar?

Estimate how loss rates have moved with an indicator in the past, then apply that sensitivity to the difference between the forecast and the historical average.

Is a forward-looking adjustment always required?

IFRS 9 requires forecasts to be considered; if the outlook is similar to the historical period, no adjustment may be needed, but the conclusion must be supported.

What if a company cannot find a statistical relationship?

It can use qualitative, evidenced judgement, such as recent payment behaviour and industry information.

Sources

Every fee, date and rule on this page was taken from these official and primary sources.

  1. IFRS Foundation: IFRS 9 Financial Instruments

Rules and fees change. If you are reading this long after October 4, 2026, confirm the figures with the source before you rely on them.

ECLCalculating historical loss ratesECLSegmenting receivables for a provision matrixECLForward-looking information in ECL

More in ECL

TopicAll 40 guides in ECL

This guide is general information. It is not tax or legal advice for your situation.