One price index, three ways of reading it, and a policy rate that depends on which one you pick.

By Kanisa Magazine · March 2025

In the third week of March 2025, the same set of Congolese prices could be described as rising at 10.4 per cent, at 9.9 per cent, or at 2.2 per cent, depending on which clock you used. A Congolese newspaper headline had even offered 6.2 per cent for the end of 2024, while the official statistics office had 11.7 per cent. None of these numbers is wrong. They answer different questions, and the Banque Centrale du Congo’s policy rate is tied to one of them.

That makes the way the evidence is built a commercial matter. Anyone pricing a contract in francs, setting a wage review or judging whether a 25 per cent policy rate is tight needs to know which reading is being used, and who produced it.

One index, three clocks

The same weekly index yields a year-on-year rate, an annualised rate and a year-to-date rate.

The BCC’s statistical digest of 26 March, covering the week to 21 March, reprints the national household price index from the Institut national de la statistique, with December 2011 set at 100. In the third week of March it stood at 474.57. Measured against the same week a year earlier, that is 10.4 per cent. Measured against the last week of December, it is 2.2 per cent, and the BCC’s annualised rate, which projects the cumulative increase to the end of the year, comes to 9.9 per cent.

The Kinshasa index, published alongside it, stood at 1,792.10 and reads slightly higher: 11.0 per cent year on year, 2.2 per cent since December and 9.8 per cent annualised. The two indices cover different baskets and different markets, so a gap of 0.6 points is unsurprising.

Why last year’s data moves this year’s rate

A year-on-year rate changes when old weeks drop out of the comparison, not only when new prices arrive.

The national year-on-year rate was 20.7 per cent in the last week of June 2024. By the second week of July it was 15.3 per cent, after 19.6 per cent a week earlier. Weekly price increases that month were 0.3 or 0.4 per cent, which cannot produce a fall of that size by themselves. The drop must come from a large weekly rise a year earlier leaving the comparison window.

The digest shows the pattern settling after that. In the third week of March the national index rose 0.147 per cent on the week. Repeated for a full year, that pace compounds to about 7.9 per cent, below every annual reading above, which tells a careful reader that the rate is more likely to keep falling than to rebound if the weekly rhythm holds. The national year-on-year rate eased from 15.2 per cent in August to 11.7 per cent in the last week of December, and to 10.4 per cent by the third week of March. A reader who quotes a single week without noting the base is quoting a statistic that has a large mechanical component.

Which number the policy rate follows

The BCC states that its policy rate is set on a positivity principle against year-on-year inflation.

The digest’s methodological notes say that the policy rate is calculated on the principle of positivity relative to year-on-year inflation, and that the overnight facility sits one percentage point above it. With the rate held at 25 per cent, the Monetary Policy Committee reported on 30 January 2025 under Governor Malangu Kabedi-Mbuyi that it was keeping it there, as Financial Afrik reported. Set against the national reading of 10.4 per cent, the real policy rate is about 14.6 points.

That is a large positive number, and it is what the rule in the notes would predict from a year-on-year series that had halved in twelve months. A trader or lender who instead used the annualised figure would see 15 points, and one using the year-to-date figure would conclude something quite different. The choice of clock changes the story about monetary tightness.

Two headlines for one year

Press versions of the 2024 outturn differed by five points, and the primary series settles the argument.

In the days after the January committee meeting, the news agency ACP reported 11.7 per cent inflation for 2024, against 23.8 per cent for 2023. The same period carried an Infos27 headline speaking of inflation reduced to 6.2 per cent. The BCC digest’s own table of INS data gives 23.8 per cent for 2023 and 11.7 per cent for 2024, with annual-average inflation at 10.9 per cent in 2024.

The 11.7 per cent figure is therefore the one that matches the published national index. A reader cannot tell from the headline alone what the 6.2 per cent measures, and a diligence file should say so rather than quote it.

Putting the evidence to work

Name the index, the base and the clock in every franc-denominated contract.

For a business, the lesson is practical. A franc price-adjustment clause that says only “inflation” leaves room for a dispute between 10.4, 9.9 and 11.0 per cent. A clause that names the INS national household index, the December 2011 base and a year-on-year comparison removes that room. The same discipline applies to wage reviews, rent escalators and loan covenants, and to any dollar contract that converts into francs at a stated rate. The BCC’s indicative rate on 21 March was CDF 2,845 to the dollar, and its net reserves series stood at US$6.25 billion, about two months of imports, so the exchange-rate leg of such a contract has been steadier than the price leg.

The national year-on-year rate eased from 11.1 per cent in the last week of January to 10.4 per cent in the third week of March, and the next BCC digest will show whether that slide continues. Treat the headline as an invitation to check the series.