As part of our ongoing market analysis, we monitor liquidity across a range of UK markets, particularly the UK cash gilt market. The healthy functioning of the gilt market matters because of its importance to UK financial and monetary stability. To monitor gilt market liquidity, we combine insights from market and transaction data with market intelligence (MI) to give us a comprehensive picture.
In April 2025, there was a period of heightened global volatility and consequently, liquidity temporarily worsened. In this post we show that overall, the level of liquidity in gilts was commensurate with the degree of volatility. In addition, liquidity deteriorated significantly less than other stress episodes in recent UK history – namely the ‘dash for cash’ in 2020, and the liability-driven investment (LDI) episode in 2022.
Why do we monitor gilt market liquidity?
Governments across the world borrow money to finance spending. In the UK, the government primarily does this by issuing bonds called gilt-edged securities, known as gilts. There are currently over £2 trillion gilts in issue. In common with other sovereign debt markets, gilts are tradable securities, meaning investors can buy and sell them in the secondary market.
The gilt market is fundamental to the broader UK financial system and real economy given its role in financing government activity; as a benchmark for the pricing of other financial instruments, such as corporate bonds or loans to households and businesses; and by serving as a safe and liquid asset for use as collateral or in capital regulatory requirements. Given the size of the market and their wide range of uses, a material deterioration in gilt market liquidity could impact monetary transmission and financial stability in the UK.
How do we monitor gilt market liquidity?
We take a holistic approach to monitoring gilt market liquidity, combining data insights with MI. Quantitatively, liquidity can be monitored using a mixture of market data based metrics and transaction data based metrics. Market data are observed in real time, such as bid-offer spreads and yield curve noise, providing timely assessments of gilt market liquidity (Table A for more detail). Transaction data metrics are available with a short lag of a few days, such as Amihud, which measures the price impact per volume traded.
We also look at other indicators which are based on closely related markets, such as the long gilt futures market, which is the only actively traded UK bond future. Gilt futures are traded on an electronic orderbook, meaning we can track how liquidity evolves from one second to the next. This is helpful for understanding how liquidity responds to particular events.footnote [1] Here, we look at measures such as order book depth, the number of contracts ready to be traded at the most competitive prices, and number of events, the number of times contracts are added, modified, cancelled or executed.
Table A: Table of a subset of gilt market liquidity monitoring metrics
Variables | Explanation | Market or transaction data |
|---|---|---|
Bid-offer spread | This is the difference between the highest price a buyer is willing to pay and the lowest price a seller is willing to accept. A wider spread implies worse liquidity. | Market |
Yield curve noise | This measures deviations of yields from a fitted yield curve. It looks to capture market participants’ willingness to arbitrage dislocations in yields. When yields are further away from a fitted curve, this can suggest lower liquidity. Although, this can also be driven by technical factors (such as how gilts are taxed). | Market |
Amihud (a) (price impact) | This estimates how much a trade of a given size causes the price to move. When a trade of a given size causes a larger move in price, it indicates worse liquidity. | Transaction |
Roll measure (b) | This measures the effective bid-offer (c) spread from transaction data. When this measure is higher, it implies higher implicit trading costs, and thus lower liquidity. | Transaction |
Price dispersion (d) | This gauges how much traded prices deviate from expected market valuation, which is based on the volume-weighted average price. When the deviations are larger, it indicates liquidity is lower. | Transaction |
Footnotes
- (a) Amihud (2002).
- (b) Roll (1984).
- (c) The effective bid-offer spread refers to the actual cost incurred at the time of the trade – whereas the observed bid-offer spread (ie the measure that uses market data) refers to the best offer (sell) price and the best bid (buy) price quoted in the market at any given time.
- (d) Jankowitsch et al (2011).
While these measures of liquidity are important, judging market functioning requires consideration of a broader range of measures. This means we also look at the speed and magnitude of price moves, as well as the degree of synchronicity of movements in related financial instruments. If an instrument is moving differently to instruments with which it usually comoves, this may indicate idiosyncratic factors (which may or may not be concerning) are causing the move. We pay careful attention to how developments in all these areas compare to historic correlations.
To enhance this holistic approach, our quantitative monitoring is always done alongside gathering and analysing qualitative MI. We gather MI through extensive discussions with financial market participants across both buy and sell-sides who are active in the gilt market, which provide us with timely qualitative information on market conditions. These approaches are mutually supportive.
Quantitative analytics form an important part of our ability to test what we hear from market participants. Similarly, interactions with – and feedback from – market participants allow us to put important context and nuance around market moves and data and can help explain the ‘why’. While both approaches are powerful in their own right, they are most valuable when used in conjunction.footnote [2]
How did liquidity during the period of volatility in April 2025 compare with the ‘dash for cash’ in 2020 and the LDI episode in 2022?
To get an aggregate overview of liquidity conditions, we have produced a composite index based on a subset of quantitative measures. Individually each of these measures alone are imperfect measures of liquidity. To account for this, our Gilt Market Illiquidity Index is constructed as the first principal component of the subset of indicators. It retains the shared variation among the indicators, which are thought to be driven by liquidity conditions, and discards idiosyncratic variation, which reflect movements that are not driven by liquidity. The factor loadings (ie the weight placed on the different components) vary over time, which means if one metric becomes less correlated with other measures, perhaps because it tracks liquidity less closely, it will have a smaller influence on the overall index.
In early April, following tariff announcements in the US, there was a period of large moves in sovereign bond markets, including the gilt market. We compare the moves in this period with two previous episodes of volatility/illiquidity:
- the dash for cash (DfC) in 2020: where economic disruptions associated with the Covid pandemic sparked a surge in the global demand for cash, leading to investors rapidly selling sovereign bonds and thus increasing bond yields; and
- the LDI episode in Autumn 2022: when the UK gilt market exhibited extreme volatility, leading to a fire sale of gilts by LDI funds and severe illiquidity in the gilt market.footnote [3]
Throughout these periods, we have used our quantitative and qualitative tools to help us understand how gilt market liquidity evolved.
Our Gilt Market Illiquidity Index (Chart 1) – where a lower level implies greater liquidity – is an effective way to compare episodes of gilt market stress.
Chart 1: Gilt Market Illiquidity Index
Footnotes
- Sources: Bloomberg Finance LP, MiFID II and Bank calculations.
The index suggests that there was a modest worsening in gilt market liquidity in early April – which can be seen in the small uptick in the shaded area in 2025. However, it did not approach levels seen during the DfC and LDI episode – as seen in the other shaded areas. In fact, liquidity remained within the average historical range observed since 2020. The deterioration in liquidity was temporary, and our index returned to pre-April levels several days later.
It is not a surprise that liquidity fell both in the gilt market and other government bond markets globally over this period. Periods of heightened uncertainty can lead to divergent market expectations for the path of future asset prices. In turn, prices fluctuate more dramatically as market participants process new information and develop their perceptions of risk – leading to increased volatility. In general, as volatility increases, liquidity conditions tend to worsen. This negative relationship is well-documented. When volatility is higher, market makers might act more cautiously to reduce their exposure to the increased risk of holding assets that might quickly change in value. They respond to this increased risk by, for example, widening bid-offer spreads or trading in smaller quantities. Chart 2 illustrates this negative relationship for the UK. It plots gilt market volatility against illiquidity (as measured by our Gilt Market Illiquidity Index) from 2018 to 2025.
Chart 2: Volatility and illiquidity in the gilt market
Footnotes
- Sources: Bloomberg Finance LP, MiFID II and Bank calculations.
Three episodes are highlighted to show days during the DfC (purple), the LDI crisis (gold) and the recent period of volatility in April (orange). In this context, the combination of volatility and liquidity during April were in line with historic patterns for the UK gilt market since 2018. Overall, this evidence aligned with our MI – which suggested that the level of liquidity had declined, though to a lesser extent than in prior episodes, and was commensurate with the rate of volatility. The volatility in April was less heightened and for a shorter period of time than the DfC and LDI episodes, which aligns with the smaller worsening in liquidity.
We will continue to monitor the market using our quantitative and qualitative tools.
In summary, this post has shown that: the deterioration in gilt market liquidity in early April 2025 was temporary, with liquidity returning to pre-April levels shortly after; the level of illiquidity was commensurate with the level of volatility at the time; and the spike in illiquidity was significantly less than during the DfC in 2020 and the LDI episode in 2022. This event study showcases how our quantitative data analysis works in harmony with our qualitative MI to effectively monitor the gilt market.
Given the environment of heightened global uncertainty, we will continue to closely monitor and analyse activity in financial markets vital to the UK’s monetary and financial stability, including the gilt market.
This post was prepared with the help of Eleanor Kantor and Joel Mundy.
Share your thoughts with us at BankOverground@bankofengland.co.uk
Futures under stress: how did gilt futures behave in the LDI crisis?.
The sequence of events is covered in detail in other Bank of England publications.