Filter insights on this page
Repo rates are expressed relative to SONIA, and the chart below displays the average repo rates that we have achieved over the past four quarters for three, six, nine and 12-month repos, shown as a spread to average SONIA levels at the time. The volatility and market uncertainty that resulted from the mini-Budget also weighed upon funding markets, particularly for shorter dated trades as can be seen from the achieved spreads below. Note that during the fourth quarter of 2022 no repos were traded with a 12m tenor so the chart reflects the previous quarter’s value.
Capabilities
Media type
Themes
The secondary impact of the mini-Budget crisis centred around collateral and the velocity of movement; rather than a lack of balance sheet for repo funding (a la March 2020). Yet, the difficulties around collateral substitutions and settlements did in many cases prompt a review by individual banks’ credit officers, resulting in a temporary reduction or hiatus in repo balance sheet provision in some cases. Once these reviews were completed balance sheet availability opened up again – some with the addition of haircuts to provide additional protection to the bank. Of course, the momentous lack of certainty in the future path of interest rates also impacted the typical repo spread to SONIA as trading a fixed rate forced the banks to take a conservative view on where yields could reach.
Regions
In the first quarter of 2023 both market participants and regulators continued to digest the ramifications of the gilt crisis of 2022 and what that could mean for the mores of leverage and oversight for the LDI market.
Whilst regulators were undertaking their analysis the LDI market also pre-emptively considered their appropriate resilience which resulted in a reduction of leverage across the street. As a consequence, January was pretty quiet in the LDI space as pension funds reviewed their situation with particular reference to how the events of September/October effected their funding ratio. For many of those who had access to liquid assets and were able to support their hedging requirements, the higher yields accelerated their flightpath to buy-out/in. In turn, February was dominated by buy-out news and activity with the largest ever deal announced as PIC competed the buy-in for the £6.5bn RSA UK pension scheme. As more schemes turn their eyes towards their end-game, much focus has been upon the resource constraints at buy-out providers, with queues likely to last multiple years for smaller schemes as insurers concentrate on ‘bang-for-buck’. However, in the meantime those schemes who do intend to target buy-out can focus on de-risking in preparation.
Globally the expectations and activities of central banks maintained their dominance over markets. The rhetoric at the start of the quarter pointed towards a slowdown or plateau of the hiking cycle and even indicated the potential to commence reducing rates for a more accommodative policy. The recent tightening of monetary policy and the interaction with some of the rules governing smaller US banks delivered a jolt to the world as runs on Signature Bank and Silicon Valley Bank resulted in their collapse.  The main issue centred around their held-to-maturity book which had been heavily invested in US Treasuries when rates were at rock-bottom, and since the rate hiking cycle were now valued much lower, yet were not required to be marked-to-market unless sold to fund withdrawals). In the modern equivalent of a bank run, these flash runs can be much harder to manage as withdrawals can occur at a click of a button rather than requiring the wheelbarrow queueing option. Indeed circa $42billion was pulled out of SVB on the 9th March precipitating the bank’s collapse. As painful memories of the 2008 financial crisis abounded, other vulnerable banks were targeted by investor flight, namely Credit Suisse which was ultimately bought (extremely cheaply) by its local rival UBS in a deal masterminded by the Swiss government.Â
The market’s view of where long-term rates could move to in the future is encapsulated in forward rates. The chart below shows where the six-month SONIA swap rate is currently (spot) and at various forward rates out to five years. As can be seen from the chart below, rate expectations have risen marginally at the front end but with a lower peak expected in the next quarter or so. One-year forward rate expectations have dropped by 0.43% within the last three months.
Figure 1: Six month SONIA rate
Source: Barclays Live, as at 31st MarchÂ
Repo rates are expressed relative to SONIA, and the chart below displays the average repo rates that we have achieved over the past four quarters for three, six, nine and 12-month repos, shown as a spread to average SONIA levels at the time. Happily, volatility was much reduced from the previous quarter returning achieved repo rates to levels similar to the second quarter of 2022, albeit with a little more traditional term structure at the longer tenors.
Figure 2: Spread to SONIA
Source: Columbia Threadneedle Investments, as at 31st March 2023
Whilst the future path of interest rates has become more digital (increase, decrease, stay the same) the range of possibilities has narrowed as markets anticipate the peak in rates in this cycle. However differing opinions as to where Base Rate could be in 12 months’ time contributed to the higher spread to SONIA seen for longer maturities. The lower volatility seen in the first quarter of 2023 resulted in a more consistent measurement of the relative repo cost to SONIA. Despite the continuation of Quantitative Tightening, liquidity in repo funding is still ample, particularly as most pension funds are using less leverage and thus their total funding requirement has diminished. Also, despite QT, much of the sub-10-year part of the nominal gilt curve remains in high demand, locked up on the Bank of England’s balance sheet – i.e. trading ‘special’. Columbia Threadneedle preferentially looks to repo these special bonds to receive a material saving on repo funding costs; this quarter a typical repo spread to SONIA for one of these bonds in high demand was around -0.41%.
Another consequence of the gilt crisis has been the adjustment of expectations for volatility or value at risk assumptions, in particular as it pertains to central clearing. This has resulted in extremely high initial margin requirements specifically for cleared repo; (note that swap-based initial margin requirements have also increased but not to the same extent). The upshot is that bilateral repo is far more attractive both from a cost and a collateral efficiency perspective. Therefore, since the last quarter of 2022 volumes have moved away from alternatives into traditional bilateral repo.
Repo funding generally remains cheaper at this moment for creating leveraged exposure to gilts over the lifetime than the equivalent total return swap (TRS) and so continues to be used within our LDI portfolios and has the advantage of using bond and cash collateral or even credit collateral. However, pricing for total return swaps can be very bond specific and, where the bank counterparty can obtain an exact netted position, the rate can be extremely competitive. TRS can be longer dated, with maturities ranging from one to three years and even five years, as compared to repo which typically vary in term from one to 12 months. Hence, TRS can be beneficial for locking in funding costs for longer and for minimising the roll risk associated with shorter-term repo contracts. On the other hand, repo facilitates tactical portfolio adjustments more easily and tends to be slightly cheaper. We ensure portfolios have access to both repo and TRS for leveraged gilt funding, so we can strike the right balance between cost, flexibility, and minimisation of roll risk. It is essential to maintain a range of counterparties to manage the funding requirements of a pension fund. We now have legal documentation in place with 22 counterparties for GMRA (Global Master Repo Agreement) and 23 counterparties for ISDA (International Swaps and Derivatives Association) and more are being negotiated.
Following the gilt crisis last year, we are seeing interest from clients in credit repo and appetite from some banks to support the same. Credit repo allows portfolios with directly held credit to raise cash to support hedging without selling their credit, once their gilt positions are depleted. Pricing is highly bank and bond dependent and can be anything from 10-50bps more than the cost of a traditional gilt repo, rising to 65bps more in a crisis. This means that credit repo should be thought of as a short-term contingency solution rather than a long-term funding tool. However, it is a beneficial addition to the toolbox and something we are putting in place for relevant portfolios.
Indicative current pricing shows leverage via gilt TRS for a six-month tenor is very bank dependent and can be either 0.03% wider than repo or a similar amount tighter – this typically depends on the bank’s view of the repo market. Another way to obtain leverage in a portfolio is to leverage the equity holdings via an equity total return swap. An equity TRS on the FTSE 100 (where the client receives the equity returns) would indicatively price around 0.11% higher than the repo (also as a spread to six-month SONIA). Clearly, this pricing can vary considerably from bank to bank and at different times due to positioning, which gives the potential for opportunistic diversification of leverage.Â
SONIA – Sterling Overnight Index Average
