The Bank of England has done some timely and truly eye-opening research into the resilience of corporate bond markets. The research is contained in the Bank of England Financial Stability Paper No.42 and is titled “Simulating stress across the financial system: the resilience of corporate bond markets and the role of investment funds” by Yuliya Baranova, Jamie Coen, Pippa Lowe, Joseph Noss and Laura Silvestri.
The starting point of the analysis is to revisit the Global Financial Crisis (GFC) which saw $300 billion of related to subprime mortgages amplified to well over $2.5 trillion of write-downs across the global financial system as a whole. One of the problems was that the system was structured in a way that did not absorb economic shocks, but amplified them. The amplification came via a feedback loop. As the crisis unfolded, fears about credit worthiness of banks led to the collapse of interbank lending. Weaker banks had their funding withdrawn, which led to a downward spiral of asset sales and the strangling of credit in the broader economy.
The paper notes that, since then progress has been made and the Bank of England’s stress tests now include the feedback loop created by interbank loans.
Indeed, the 2016 test showed that the potential for solvency problems to spread between UK banks through this channel has “fallen dramatically” since the crisis. Furthermore, interbank lending has been cut back and is more often secured against collateral.
The report cautions that other feedback loops might be present, especially since banks only account for about half of the UK financial system. Indeed, a key objective for regulators is to assess how the non-bank part of the system – termed “market-based finance” in the paper, responds to economic shocks. In particular, could the non-bank system, which trades “market-based finance” (principally bonds), amplify shocks in a similar way to the banking system during the last crisis? The report characterises market-based finance and the related risks as follows.
The system of market-based finance includes, among other parts, investment funds, dealers, insurance companies, pension funds and sovereign wealth funds. It supports the extension of credit and transfer of risks through markets rather than banks. It has expanded rapidly since the crisis. At the global level, assets held by non-bank financial intermediaries increased by more than a third since the financial crisis. The potential spillover effects in market-based finance centre on ‘fire sales’ of assets, which affect prices of financial assets and functioning of markets. Participants in this part of the system can face incentives, or be forced into, sudden asset sales.
The report sees the potential for another dangerous feedback loop developing from falling asset prices which lead to declines in net worth, prompting a withdrawal of funding which leads to more asset sales and further falls in prices. They are hardly reinventing the wheel here and what they’re really describing is the evidence that investors often behave pro-cyclically. It raises the valid concern that pro-cyclical behaviour is most dangerous in less liquid assets with short-notice redemption – the classic liquidity mismatch. The post-Brexit problem in 2016 in UK commercial property funds was a great example.
These dynamics were illustrated clearly in 2016 in funds investing in UK commercial property. With the property market in hiatus following the United Kingdom’s referendum on membership of the European Union, these open-ended funds faced redemption requests from investors concerned about the prospect of future price falls and fearing that other redemptions would force the funds to suspend. The process was self-fulfilling and many funds were forced to suspend redemptions.
The report goes on to highlight the challenges for broker-dealer liquidity and hedge funds if asset managers aggressively sell securities in a crisis. It’s obvious stuff, i.e. that broker-dealer are less able to warehouse securities and less able to provide funding to hedge funds, which might be buyers, and could become forced sellers. The BoE models what would have when one type of shock - redemptions by open-ended funds - trigger selling by the funds with spillover effects for broker-dealers and hedge funds.
The paper that follows seeks to model how the aggregate behaviour of several sectors within the system of market-based finance, including investment funds and dealers, could interact to spread and amplify stress in corporate bond markets. That focus stems from the growing importance of bond markets to the financing of the economy, alongside the rapid growth in holdings of such bonds in fund structures. It does not focus on individual companies; the analysis is conducted at a sector level. It is not concerned with the capacity of the sectors to absorb losses.
Basically, the model estimates the sensitivity of investment grade corporate bonds yields in Europe if funds sell the equivalent of 1% of their total assets on a weekly basis – which was similar run rate to the redemptions in October 2008 (4.2% over the month – see below). Since then, however, broker-dealer capacity has contracted and investment grade issue issuance risen sharply. Importantly, it also addresses the scale of redemptions which might overwhelm the ability of broker-dealers and hedge funds to absorb the selling. The model assumes that there is a shock leading to an initial round of redemptions which prompts investment funds to make asset sales. Broker-dealers require lower prices to compensate them for absorbing the selling which leads to a second round of redemptions and selling. After that, further selling “breaks” the market and leads to dislocated prices on the downside.
The paper explains the market-breaking points as follows.
The level of redemptions at which the second-round price impact line ends is where dealers reach the limit of their capacity to absorb those asset sales by funds not purchased by hedge funds. We assume that market liquidity is tested at this point and refer to it as the market-breaking point. Transactions could still occur beyond this point — for example, if a dealer can immediately match a buyer and seller or if it sells other assets to purchase corporate bonds — but are assumed to take place at highly dislocated prices.
The BoE paper estimates that a weekly level of redemptions from funds equivalent to 1% of their assets would increase investment grade corporate bond yields by 40 basis points. However…this is the key…it estimates that initial redemptions equivalent to only 1.3% of assets on a weekly basis would be “needed to overwhelm the capacity of dealers to absorb those sales, resulting in market dysfunction”, i.e. the market-breaking point. It describes this as an “unlikely but not impossible event.”
We disagree, we are in a far bigger bubble than 2007-08.