Germany needs more than a plan

With inflation and growth picking back up, we expect the Fed to deliver a final rate hike at its November meeting, after pausing in September. The CPI report released on Wednesday showed that inflation picked up to +3.7% y/y in August, a step up driven by higher gasoline prices. Furthermore, the pace of core price increases stepped up, to +0.3% m/m (after +0.2% in the two previous months). The further rise of the WTI price at the beginning of September should delay the pullback in consumer prices. Against this backdrop, the incoming economic data strengthen the expectation that Q3 GDP growth picked up pace through the summer months. Crucially, two powerful channels through which monetary policy is supposed influence aggregate demand
– house prices and bank credit – have picked back up. If sustained, this could prolong above-trend growth and risk re-igniting underlying price pressures. Therefore, on balance, we think that the FOMC participants will press ahead with a 25bps rate hike in November after pausing (as widely expected) at next week’s meeting.
We acknowledge, however, the large uncertainty surrounding this call: in order to further their hawkish stance, FOMC participants could also emphasize their ‘’higher for longer’’ narrative instead of hiking rates

Elevated interest rates are sending Fed’s financial losses into the red. The potential medium-term implications of the Fed making increasingly large financial operating losses are drawing increasing attention. This is the first time since 1915 that the Fed is making losses. Over the past 20 years, it had remitted an average of USD63bn annually to the Treasury. But starting in Q4 2022, the Fed started to make increasingly large operating (as well as unrealized) losses (Figure 1) as the mismatch between what it earns on its assets and what its pays on its liabilities is widening. Indeed, the Fed owns a very large portfolio of securities (mortgage-backed securities and Treasuries) that yield a fixed interest rate. As all of this portfolio – currently about USD7.3trn – was purchased prior to the rise in interest rates in 2022, it yields a very low effective interest rate, a legacy of the pre-Covid era, which we estimate at around 2.27 % in Q2 2023. Conversely, the Fed has to pay a variable, and sharply rising, interest rate on its liabilities (banks’ reserve balances and repurchase agreements): we estimate the effective interest rate on its liabilities to have been around 5.03% in Q2 2023, up from only 0.81% in Q2 2022. The cumulative operating losses stood at USD74bn in Q2 2023 (Figure 1).

 

Does it matter if a central bank makes losses? The Fed’s cumulated USD74bn losses have already largely outstripped the central bank’s USD42bn in capital. If the Fed was a commercial bank, it would be technically insolvent. But it does not really matter for a central bank to operate with negative equity: For instance, the central banks of Chile or Czechia have carried negative equity for a long time without being hampered in their ability to fulfill their objectives. A central bank’s credibility depends first and foremost on its ability to meet its mandate and sometimes making losses ‘’can be the price to pay for achieving these aims’’ (BIS, 2023).

Large, sustained losses can be problematic in the medium term as they may erode the public’s trust in the institution. However, current times may be different in regard to the size of current and future losses the Fed (and other central banks) are and will be making. Under our assumptions that the policy rate settles at 4.25% (upper band) in end 2024 and 3.25% in end 2025, we estimate that the Fed’s cumulative losses will keep increasing until the middle of 2025 to around USD186bn, or 2.8% of the Fed’s assets, before slowly reducing thereafter as the Fed starts to make profits again (Figure 2). Such large losses could end up drawing the public’s attention and start to erode its trust in the Fed’s ability to fulfill its mandate, including that of maintaining price stability. Indeed, the Fed creates bank reserves to cover operating losses, that is central bank money, which are in principle inflationary. Large losses can thus go against the Fed’s goal of price stability ultimately. 

In the near term, the Fed is nudged to continue with quantitative tightening (QT), but in 2024 it will be incentivized to cut rates sooner rather than later. For now, as inflation is still elevated, it would be foolish for the Fed to start cutting rates. The most sensible way for the Fed to remain credible in its fight against high inflation while minimizing losses, is to continue with passive QT, i.e. letting its holdings of securities roll off as they mature. Passive QT allows the Fed to reduce the bank reserves as securities now need to be funded by the private sector. Lower bank reserves, in turn, mean lower interest to pay for the Fed. However, at the current pace (about USD100bn of securities being rolled off every month), the Fed may have to stop QT (and stabilize its balance sheet) by the end of Q2 2024. Indeed, we estimate that by then bank-reserves balances will have reached about 7% of GDP (Figure 3), a level below which volatility in the repo market may arise due to a shortage of liquidity, as the episode of September 2019 showed. Lowering interest rates will then become the only realistic option for the Fed to stop making losses. If, as we expect, inflation does normalize in 2024 (approaching the 2% target), we think that keeping interest rates too high for the Fed may backfire. Under a scenario of ‘’higher rates for longer’’ where the Fed does not pivot in 2024 (interest rates remain at about 5.5% by end 2024) and starts cutting rates slowly in 2025 (down to 4.5% by end 2025), we estimate that the US central bank would continue to accumulate losses (Figure 3). In other words, if inflation does subside next year, there is a strong argument for the Fed to pivot sooner rather than later, otherwise predictable large and accumulating losses could hurt its credibility to fulfill its objectives.   

We expect one 25bps hike from the Bank of England at its next meeting on 21 September, which will increase its key interest rates to 5.5%, followed by a peak at 6% in December 2023. We expect rate cuts equivalent to a total of -125bps will only begin after H2 2024, which will push key rates to 4.75%. We believe inflation will remain high at slightly below 5% by December 2023, with continued elevated pressures from the services side. We do not expect inflation to fall below 3% in 2024 (Figure 4). 

There are three reasons why the Bank of England’s rates will peak at 6%. First, the tightness of the job market and the risk of a wage-price loop. The rebalancing of the UK labor market is ongoing but remains very slow. Job vacancies have fallen by -24% since their peak, but remain above trend (unlike in Germany, for example, Figure 5).

The fall in employment picked up speed (-207K 3m change in July, three times more compared to June) and the unemployment rate remains very low (4.3% in July, +0.1pp from one month earlier). However, despite labor shortages cooling down, there seems to be no turning point in wage growth which still runs above 8% 3m y/y in services and manufacturing, and will probably stay well above the Bank of England’s forecasts of 6.9% in September on average. Total real wage growth has been positive since June, and since April in the services sector (Figure 6).

This is clearly a trend to monitor in the coming months as a wage-price loop could become a real risk if real wage growth is positive during three out of the next four quarters, i.e. until mid-2024. The IMF has identified two such episodes in the past in the UK (2003, 2016), much less than in the US, which has seen six such episodes (1973, 1978, 1987, 1996, 2000 and 2017). Belgium and Germany, where wage growth is also currently very dynamic, have seen four (1999, 2005, 2010 and 2016) and three (1989, 2010 and 2017), respectively. As 2024 is a political year, we expect the minimum wage to be raised by at least another 7% after 9.7% this year, against inflation forecasts of 3.5% and 7% respectively, which means an acceleration in terms of the real wage growth. Total real wage growth is expected to reach a cumulative +3.9pps, compared to -3.1% in 2022 (Figure 7). 

Second, the corporate price outlook has stabilized at high levels and corporates margins are still growing. Companies’ cost expectations have receded but they still remain higher than 2019 (Figure 8 a). This can be partly explained by higher labor costs. At the same time, expectations for selling prices are stable at high levels in B2B services, retail goods and services, while they adjusted downwards in the manufacturing sector to their lowest level since end-2021 (Figure 8 b). Corporates’ EBITDA margins grew by +1pp of gross value added between Q2 2023 and Q2 2022, a sign that UK firms did not take the cost increases on their margins despite the acceleration in unit labor costs. This is clearly in contrast to the Eurozone and the US where margins have started to fall. Additionally, companies’ cash buffers have been decreasing (-15% excess cash in July 2023 compared to end-2022). 

Finally, the economic outlook suggests lower recession risks and negative real interest rates until year-end. While the external environment has been a drag on activity in 2023, with exports decreasing by -3.1% y/y, domestic demand, and in particular the consumer, has been resilient. Despite the fall in GDP in July, we do not see a recessionary trend ongoing as the -0.5% m/m drop was probably driven by the number of days of strikes (almost twice as many working days lost) and unusually wet weather that pushed construction output into contraction. On one hand, UK households entered the tightening cycle with a large stock of savings, which currently stand at 4% of GDP. On the liabilities side, the bulk of outstanding mortgages is fixed for either two or five years. This is slowing the rise in effective average mortgage rates on the outstanding stock of debt and is making a difference of at least 200bps on interest rates on the flow and the stock of loans. The negative interest rates (-1.8% as of July) are expected to prevail until year-end, even though they already turned positive in the US back in April. This gives some additional time in terms of debt sustainability, and provides the Bank of England with some incentive to continue to increase interest rates to further cool down demand, notably in the services sector. While we expect the UK to avoid a recession, GDP growth will remain low at +0.3% and +0.6% in 2023 and 2024, respectively. 

The results of our annual Allianz Pulse survey on the assessment of the economic situation reveal little that is surprising: the mood is bad, the outlook bleak. What is striking, however, is that the German participants in the survey are starting to pick up the pessimism of the French and Italian participants (Figure 9). While "German optimism" still stood out in 2019 and 2021, this year the negative voices clearly predominate: the net percentage stands at -25.1%. This means that German participants assess the current economic situation as even worse than in 2020, when the survey was conducted in the shadow of the first major Covid-19 wave, and in 2022, when the shock of the Russian war of aggression on Ukraine was still fresh. Significantly, this is not the case in either France or Italy, where survey participants are currently not quite as pessimistic (even if the absolute level is still lower than in Germany). It seems that German participants are only realizing with some delay that the geo-economic framework is currently shifting to Germany's disadvantage.

The differences between the generations are not very pronounced, even if the youngest participants (Gen-Z) are generally somewhat less anxious about the future. But even among them the pessimists clearly predominate: net percentage -22.5% vs -25.2% for the entire sample.

What also distinguishes this year's survey is the large discrepancy between the assessment of the general future outlook and that of the personal outlook. While in all years participants tended to see their own future as somewhat rosier than that of their respective countries, the gap was only as large as this year in pandemic times; in Germany, it reached an all-time high this year (Figure 10). The differences between the three countries are therefore no longer so pronounced. This also applies to the generational comparison. Although the "personal gap" is highest in the two younger generations (Gen-Z and Millennials) at around 50pps, it is also very high among the older participants at just under 30pps.