Gauging inflation impact of the Red Sea crisis

After noting that we are again ‘In Deep Ship’ last month, the West’s Operation Prosperity Guardian unfortunately has not returned the Red Sea and Suez Canal to normal conditions. Higher insurance costs and risks to staff are forcing ocean carriers to continue to avoid both (see figure 1). The regional situation remains delicate, with the risk of further military escalation.

For now, naval escorts are the only way for limited numbers of Western ships to utilize Suez, which is a high-cost exercise. The EU is readying a naval mission of three ships, dubbed Eunavfor Aspides, but this is unlikely to be a gamechanger, given the small scale of this mission compared to the density of commercial shipping in the region. As such, it seems likely that many ships will continue to take the detour around the Cape of Good Hope.

Longer travel times and arguably the opportunism of shipping companies have pushed up freight rates. In this report, we look at the potential consequences for inflation.

Maritime freight rates are exhibiting mixed trends in recent weeks. While Shanghai-Genoa and Rotterdam routes witnessed a slight easing, other routes – including those to the US – have experienced upward pressure (see figure 2).

Going forward, more carrier supply becoming available in 2024-25 could alleviate some strain. This is unless older, polluting vessels are scrapped at the same time due to the phasing in of ETS allowances in 2024[1] and the (updated) EU Ship Recycling Regulations. These regulations will be implemented according to the requirements of the Hong Kong Convention, entering into force on 26 June 2025.

Yet issues with shipping containers could do the exact opposite. Containers are being repositioned due to global trade imbalances, with knock-on effects, as experienced during Covid. The increasingly irregular arrivals of ships due to the Red Sea crisis is causing intermittent congestion at certain ports. Experts caution a comprehensive resolution to this crisis may take several quarters, and costs are likely to remain elevated, even without further geopolitical shocks in the area. Last week, the head of the International Maritime Organization warned that piracy risks along the African coast may also increase, and insurance costs would increase as a result.

Moreover, the ocean carrying capacity for key commodity cargoes such as Liquified Natural Gas (LNG) and Liquified Petroleum Gas (LPG) will not have the same positive ocean-carrier supply effect ahead. Carrying capacity remains very tight there, given the huge increase in journey times needed to go from the Middle East to Europe via South Africa.

This report does not assess the potential inflationary impact specifically caused by the likes of higher diesel prices, but it is something we are watching closely.

[1] The maritime sector must start paying for CO2 emissions from this year onward: In 2024, ETS allowances are required for 40% of the emissions. In 2025, this increases to 70% of the emissions. And starting from 2026, 100% of the emissions must be covered. ETS allowances must be submitted by September 30th of the following year. So, for emissions in 2024, the allowances need to be arranged by September 30, 2025.

Although distribution costs typically determine some 1% to 1.5% of the cost share of most goods, a doubling or tripling of certain (container) shipping costs, as seen since October last year, could still have a material impact on importers’ and producers’ costs.

If a container holds 1,000 items, and the freight rate rises from USD 1,500 to USD 4,500 as just seen, it implies each item sees a USD 3 increase in price to cover this. The inflation rate depends on the price of the good: assuming a USD 100 item, the rate is 3%; for a USD 50 item, it’s 6%, etc. In short, higher prices for transport over water could ultimately feed into higher consumer prices, even though it may be some time before they actually reach households and its extent depends on many other factors.

Indeed, whether wholesalers and retailers pass these higher costs on to households depends on many factors, such as alternatives (either in transport modality or product), inventory levels, and, more generally, the demand situation.

One of the key drivers of the surge in goods prices in 2021-22 was the disruption of global supply chains and the congestion of distribution over water and land. With in mind, we consider three approaches to gauge the Red Sea crisis’ potential impact today:

1. The historical estimated impact of supply chain disruptions by way of a regression analysis of Eurozone inflation surprises (outcome versus Bloomberg consensus) on both the NY Fed Supply Chain Pressure Index, (SCPI, see Box 1) and the aggregate index of container freight rates (WCI). This methodology uses models and/or analysts didn’t take such disruptions into account.

2. Explicitly adding the SCPI and WCI to a number of model specifications for core and/or ex-energy goods inflation to estimate its impact on overall inflation (assuming this improves its forecasting power).

The experience of 2021-22 shows that passthrough-effects to inflation were initially quite slow, but the unclogging of distribution chains actually came quicker than some pessimists had assumed. In hindsight, the ECB’s bullish estimates and assumptions actually proved fairly accurate.[2] Our next step in this second approach is to look at three scenarios for the evolution of supply chain disruptions based on the WCI and SCPI and to what extent this would translate into Eurozone HICP, using the “elasticities” found in the two approaches.

3. An input-output analysis looking at how and to what extent a shock in the distribution cost over water feeds through to other sectors, and hence product prices, taking into account potential substitution effects.

This third, alternative approach, uses only the WCI, on the assumption that this is an acceptable leading indicator of broader freight rates. The higher freight rates are then fed through an input-output table matrix.

[2] In October 2021, ECB President Lagarde said that the ECB expected supply chain bottlenecks to last until at least 2022Q1 and that it would, overall, “take a good chunk of ’22 for it to be sorted out”. By early 2023 the NY Fed supply chain pressures index had returned to earth.

Approach 1

Inflation surprises in 2021-22 were largely due to markets underestimating the impact of higher European gas prices and supply chain disruptions. The simple fact that oil rather than natural gas often featured in inflation models shows that the economists who ran them overlooked both this key input and broader supply chain disruptions. Both, to be fair, were of a much smaller scale in pre-Covid decades.

To illustrate, figure 4 shows that core inflation forecast errors clearly correlate with the SCPI, albeit with a lag.

In our analysis below, we use the headline inflation surprises, as we believe they are a more reliable gauge (more respondents, timelier publication) for surprises.

Our first method finds that each point (permanent) increase in the SCPI leads to a 2.7 percentage point increase in Eurozone consumer prices (with a standard error of 0.4 percentage point), when we control for several other variables.

In an alternative specification where we use the WCI instead of the SCPI, we estimate that a permanent increase in freight rates of USD 1,500/container leads to a 2 percentage point increase in consumer prices (with a standard error of 0.3 percentage point).

In both cases, we find a considerable lag in the passthrough to HICP (forecast errors) of changes in either the SCPI or WCI. It takes at least several months before impact becomes visible at all, and a “maximum impact” is only reached after some 12 to 24 months. See Appendix A-1 for more detail.

Approach 2

The second approach is to explicitly “insert” the WCI or SCPI into the regular models we use for projecting core inflation. We test this with two variants, the first with Eurozone core inflation, the second using ex-energy goods prices. However, we find very similar results for both, and so we only report on the first variant.

In this analysis, we find that a permanent increase in the WCI of USD 1500 leads to a 1.4 percentage point increase in core HICP (0.4 percentage point standard error); a permanent 1 point increase in the SCPI leads to a 1.9 percentage point increase in core inflation with a 0.3 percentage point standard error.

These “elasticities” are somewhat smaller than those found in the first analysis (see figure 5 and 6), in part because we only focus on core inflation.[3] Here, too, it takes several months before households start to feel any impact. See Appendix A-2 for more detail.

[3] We assume here that the impact on non-core components is similar, but it could be lower or higher (for example due to energy prices being affected by supply chain disruptions).

We can now entertain a number of future scenarios for the SCPI or WCI, and use the elasticities obtained above to estimate an impact range.[4] We distinguish three cases:

1. Improvement: The Suez canal reopens over the course of 2024, and problems progressively abate by the end of the year
2. Status quo: The Suez canal remains “off limits” throughout 2024, and distribution costs stay elevated until at least early 2025
3. Worsening: The Suez Canal remains “off limits” well into 2025. Problems worsen and distribution costs rise further. There is no normalization in sight by the end of 2024

Scenario A: Improvement

Scenario B: Status quo

Scenario C: Worsening

From scenario to inflation impact

In the next step, we have translated the scenarios into three stylistic paths for the SCPI (see figure 7) and WCI (not shown). We have then fed these scenarios through the models estimated in approach 1 and 2. We show the average results in figure 8. Given the uncertainties surrounding the estimates of the elasticities, we also show a low (minus 1 standard deviation) and a high variant (plus 1 standard deviation) for elasticities assumed. This gives us an impact range per scenario.

In a status quo scenario, HICP would rise by between 0.2 to 1.1 percent over a two year horizon (second column in figure 8). In a worsening Red Sea situation (third column), this could rise to 0.4% to 1.8%, which is quite significant.

[4] We assumed that the latest SCPI readings do not take full account of the current situation and its passthrough effects yet. For example, the ISM survey plays a significant role in determining this indicator and: i) it will take more time before this becomes visible, while ii) it may underestimate the problems for Europe more specifically.

In terms of likelihood, we believe scenario A is too optimistic as things stand right now. Scenario B is more likely, but scenario C is definitely possible as well. As such, this analysis shows that there is also a significant chance that Eurozone inflation will be materially impacted in 2024-25.

This also implies that the ECB will have to include this in its risk assessment. A status quo scenario may still seem manageable (in part because the current disinflationary trend is the result of thawing supply chains during 2022-23), but a worsening of the situation could even upend the current disinflationary trend.

Alternative approach: input-output analysis

An alternative approach was used by our colleagues in the Netherlands economics team. They used an input-output analysis to estimate the passthrough of a price shock in one particular sector – maritime transport – on other sectors, by taking into account substitution effects.

The idea[5] here is that a price shock in maritime transport filters through to other sectors in the economy that use it, although businesses will try to find alternatives (which will then drive up prices of these alternatives, such as air transport).

They find that a 177% price shock to maritime transport costs, in line with what we have seen so far, when sustained for the rest of the year, would eventually add around 0.8% to HICP in the Eurozone as a whole. This is closer to the lower right triangle of results in figure 8.

Their analysis also gives some indication of which sectors would be most impacted by such a rise in maritime freight costs. Unsurprisingly, this includes water, air and land transport, but also sectors such as:

[5] The prices are propagated across sectors through production functions of a sectoral producer. These functions process their intermediate input using a constant elasticity of substitution (CES) function, resulting in sectoral output along with the factor of “labor”. In reality, this factor is a combination of capital and labor, although OECD tables do not further disaggregate it. The added value is attributed to this last factor, and its income accrues to a representative consumer per household, who then utilizes it for final consumption. The weighting of sectoral output in this final consumption measure is used to calculate the Consumer Price Index (CPI). They assume a 0.3 percentage point impact on Dutch inflation, as explained here [in Dutch].

Of course, it is not only the security situation in the Red Sea or the price of container transport that affects the future impact on HICP. Here, we discuss a number of factors that may soften or amplify any impact from current supply chain disruptions.

Softer impact

One argument that can be made is that inventories are significantly higher now than during the 2021-22 episode. As the global economy was recovering from the pandemic shock, a bullwhip effect amplified distortions as businesses started ordering more than necessary. Arguably, this may not be the case this time around. German PMI surveys, for example, show that weak demand is currently outstripping higher cost of inputs, and businesses seem happy to have lower inventories.

Companies have also learned the hard way and have taken measures in recent years to strengthen the resilience of their supply chains. 2020 was the year of rate cuts and quantitative easing (QE); 2022-2023 saw significant rate hikes around the world and a gradual wind down of asset purchases. The higher interest environment leads to lower trade in investment goods and makes it more expensive to hoard inventories (and as such takes away incentives to “over-order”).

Global demand is also weaker now, which is particularly felt in the goods-producing and exporting sectors.

Bigger impact

Of course, a deterioration in the security situation in the region could make matters a lot worse, especially if other shipping routes were to be affected as well. Important to note in this respect is that Europe is more exposed to any disruptions in the Middle East due to its geographical location and trade flows, than, for example, the US. So the US-centred SCPI may underestimate the problems that European companies are facing.

Diesel-import from the Middle East and India is obviously at risk as well: Bloomberg reports that a shortage of oil tankers is a key risk given a 40-year low in large crude carrier deliveries. Although more container carrier supply should become available in 2024-25, alleviating some strain, older, polluting vessels are scrapped at the same time ahead of EU regulations.

Moreover, inventories and refinery capacity in Europe are already very low and diesel runs through the entire economy, hence feeding into the price of many goods as well as transportation services.

There is also the possibility that (global) demand recovers more rapidly. We expect Eurozone economic growth to stay weak in the short term but improve over the course of the year. Yet consumer confidence has been on the rise in recent months. If the recovery in real wages is sustained, this could also spur businesses to raise prices once again.

There has been a lot of emphasis on “real wages” since the energy shock, but what is often overlooked is that compensation policies by governments alleviated that shock quite a bit. This is illustrated in figure 12. Consumer confidence has been on the rise since late 2022.

That could also mean that, as real disposable incomes rise faster, demand may rebound more quickly than currently expected.

If consumption accelerates, bear in mind that a lot of the goods come in containers from China.

The Red Sea crisis continues. Container freight rates have dropped from their peaks on some routes, but they remain very high and cost increases have spread. If sustained, higher logistics costs could still materially affect inflation in the Eurozone. We assess the potential impact from several angles, assuming no further geopolitical shocks.

Our base case assumes a 0.5% percentage point contribution to inflation (spread over two years), but we provide a broader range of lower and higher estimates based on a number of plausible scenarios. In a worsening Red Sea situation, the impact could increase up to 1.8 percentage points, which would be quite significant – also from a monetary policy point of view.

Higher inventories, supply-chain resilience, and soft demand may dampen any impact, while a stronger-than-expected pickup in consumer spending and Europe-specific exposures may amplify this shock. It is hard to model non-linear or interaction effects with things such as energy prices, but it is a risk that cannot be dismissed entirely.