Rates: The hidden force pushing gilt yields to record highs

Global factors add to GBP term premium, but not entire story The 10Y gilt yield has hit its highest level since 2008, and whilst the near-term focus remains on oil prices and politics, the Bank of England’s balance sheet unwind poses further upside risk. And with the discussion about quantitative tightening (QT) flaring up again in the US under new Fed leadership, markets can learn from the potential impact on US rates by looking at the UK. Sterling rates are relatively sensitive to global market moves, and the spillovers from the US term premium play an especially important role. Between 2023 and today, the correlation of weekly changes in the 2s10s of the GBP curve and the US curve has been a significant 0.6. If we take the 2s10s term spread as a rough proxy of the term premium, both seem to be similar at first sight. The market optimism in US equity markets helps the US curve steepen again and also adds to the upward pressure on gilt yields. Japan’s curve shows a very different picture, and therefore we cannot simply assume a single global factor explains all term premia. For one, Japan’s economy has followed a very different path compared to other countries, experiencing stagflation for prolonged periods of time. Second, and of more importance to our analysis, is the fact that the Bank of Japan has intervened significantly more in the rates market through balance sheet policy than other central banks. Global term premia are strongly correlated but domestic factors play a role too Source: ING estimates, Macrobond Besides a global factor, the term spread is also driven by the business cycle. When a central bank hikes rates, curves tend to flatten, and vice versa. These moves seem more pronounced than purely explained by changes in the expected policy rates. As such, the 2Y rate is arguably an important ingredient for the term premium. And indeed, when looking over the past decade, the 2Y and the 2s10s show a strong negative correlation. The interesting observation here is that the term premium of the GBP curves seems to have increased disproportionately to the change in 2Y rates. This effect is also visible for the USD curve, but to a much lesser extent. When accounting for 2Y dynamics, the GBP curve recently steepened more than the USD curve Source: ING estimates, Macrobond We explore the divergence by looking at the spread differential of the 2s10s between the GBP and USD curve. Besides the 2Y spread differential, we add relative inflation to the model, which greatly improves the fit when used in conjunction with 2Y rates. Inflation differentials capture changing macro regimes and can help explain the term premium investors demand for holding longer dated bonds. Our three variable regression model (Fig 3) captures 83% of the variation in the difference between the 2s10s of the GBP and USD curve, whereby especially the residuals are of interest. We can easily identify three events that explain the deviation from our model, namely 1) QE by the Bank of England starting in 2016, 2) the Lizz Truss moment in 2022 and 3) the Fed’s tapering of QT in 2024. Outside of these periods the residual is close to zero, suggesting we are not missing any important variables. QT seems to play a part in explaining a 60bp steeper GBP 2s10s curve Whilst the Fed announced a significant tapering of QT in May 2024, the pace of QT has remained much higher in the UK, helping explain the residual in our model. The Bank of England is continuing with active sales, bringing the wind-down to a significant £70bn per year. In terms of numbers, the 2s10s GBP curve is now some 60bp steeper since this divergence in monetary policy. The Fed faced reserve constraints, forcing it to slow QT, but the Bank of England does not have this problem, allowing QT to continue. Even though shrinking the bond book has extracted reserves from the banking system, a significant uptake of the Short Term Repo (STR) and Indexed Long-Term Repo (ILTR) have managed to stabilise reserves. In effect, the market stress in US repo markets never showed up in sterling markets. Significant pickup in BoE liquidity facilities are preventing reserves from falling, allow QT to continue Source: ING estimates, Bank of England, Macrobond With money markets seemingly stable, in theory, the Bank of England can continue unwinding the gilt portfolio, albeit at the expense of higher government borrowing costs. An August report from the Bank of England suggested that QT had an impact of 15-25bp as of then on longer-dated rates. But when you also include the changes in swap spreads, our estimate of the impact from QT on gilt yields is significantly higher. Since the start of QT, the 10Y Gilt-swap spread rose more than 80bp, adding to the steeper swap curve. With more gilt supply continuing to hit the market, including significant government issuance, investors may demand higher yields to be lured in. Already we see that gilt auctions have started to rely more on foreign demand to clear since the start of QT in 2023. Before 2023, the foreign uptake of gilt syndications up to 20Y was just 10%. This increased to over 25% after the start of QT. Foreign investors tend to be more price sensitive compared to domestic players such as banks, pension funds and insurance funds. Domestic players are more likely to buy gilts for reasons like liquidity and hedging regulation. Since QT, gilt markets have relied on more price-sensitive foreign investors, helping push up 10Y gilt yields Source: ING estimates, DMO, Macrobond So while we continue to see a rising term premium for the UK, we think a dovish repricing of policy rates should start bringing down the curve. But the scope for lower 10Y gilt yields is limited. We identify 4.5% as a realistic fair value in 2027. We believe the dynamics in the UK serve as a good lesson for Fed policy too. If new Chair Kevin Warsh manages to follow up with his plans to unwind the Fed’s bond portfolio at an accelerated pace, then the US curve should see more steepening.