Risk Awards 2016
Measuring the liquidity profile of a diverse portfolio of assets is proving to be one of the great challenges of the day for the world's buy-side firms. Not only is it required by a growing number of regulators, but freak market events such as the huge intraday moves in the US Treasury market in October 2014 and the sudden spike in the Swiss franc in January 2015 have highlighted the need for enhanced analysis of liquidity conditions.
"No matter what regulators want us to report, we believe it is very important for us as a fund manager to monitor liquidity in our funds and ensure we can meet our underlying obligations. We needed a system that could assist us in creating a fund liquidity profile and show how much of the fund we would be able to liquidate under both normal and stressed market conditions," says Tonny Hald, risk manager at Nordea Asset Management in Copenhagen.
That business requirement led the firm to MSCI, whose LiquidityMetrics platform measures market liquidity across asset classes and performs liquidity stress testing. The genesis of the product dates back to 2010, when the idea was first conceived by RiskMetrics, prior to its acquisition by MSCI. Six years on, the product is now seen to be unique in the market.
"LiquidityMetrics enables us to explain liquidity in a very transparent way to staff and regulators that don't have access, because the analysis covers the real cost of trading a certain size in a certain market environment. We didn't find any other vendors capable of providing a system like this," says Hald.
Accurately capturing the liquidity of an asset or portfolio, according to MSCI, relies on two key elements: reliable models and liquidity-sensitive data. The LiquidityMetrics model is based on expected transaction cost, order size and time horizon - three variables that are brought together to create a so-called ‘liquidity surface'. The model analyses orders rather than individual transactions and assumes the user to be a liquidity-taker using professional execution channels.
"There is no consensus on a mathematical model for liquidity risk and we don't believe there can be a single liquidity scoring system," says Carlo Acerbi, executive director for risk and regulation research at MSCI. "Liquidity is not a single number, so we have modelled it as a multi-dimensional problem along transaction cost, size and time horizon of an order. The model allows users to gain a comprehensive representation of the liquidity of assets within the portfolio."
But even the most innovative model requires accurate data, and sourcing liquidity-sensitive data has proven to be a particular challenge in some asset classes. While there is a decent volume of relevant and readily available data in the equity market, it becomes much more difficult for bonds and over-the-counter derivatives, where there is no central exchange or consolidated cache of trade data.
The liquidity observatory collates data on expectations of market depth from individual points on the network, creating a very unique model that has attracted significant interest from both market participants and regulators
Carlo Acerbi, MSCI
The ability to estimate the liquidity of a bond has become one of the cornerstones of LiquidityMetrics. Given the scarcity of data, MSCI has come up with a detailed questionnaire that it sends out to participating clients, aggregating the results to generate a unique – and shifting – picture of liquidity. The initiative, known as the ‘liquidity observatory', has been in full production since May 2015.
The observatory is effectively an alliance of buy-side contributors in which front-office experts provide indications of liquidity surfaces for those segments of the market with which they have experience. Key information captured includes ‘relaxation time', which is the expected time needed to regenerate the liquidity removed by a market participant. Data is submitted on a monthly basis and participants are incentivised to take part because in return, they receive the aggregated data for free.
"The bond market is effectively a social network, in which everyone trades with their own connections without a central repository of trade and quote data, so it becomes very difficult for everyone to even observe liquidity. The liquidity observatory collates data on expectations of market depth from individual points on the network, creating a very unique model that has attracted significant interest from both market participants and regulators," says Acerbi.
At its launch in 2013, LiquidityMetrics could well have been seen as a product ahead of its time. Liquidity risk had certainly become a more widespread concern by then, and some buy-side firms were already facing new requirements through Ucits and Europe's Alternative Investment Fund Managers Directive (AIFMD), but in many cases it was superseded by other regulatory priorities.
That has now started to change, particularly following a proposal by the US Securities and Exchange Commission (SEC) in September 2015 for a comprehensive set of liquidity risk management and reporting requirements that would affect open-end funds in the US, including mutual funds and exchange-traded funds. LiquidityMetrics now has around 20 clients, of which five are contributing to the liquidity observatory.
"We initially saw more interest from clients in Europe because of Ucits and AIFMD, but there has been a much stronger impetus from the US since the SEC rules were published. Many firms now recognise the value of the methodology we are providing, which enables investment managers and risk managers to better understand the drivers of liquidity within their portfolios and meet their regulatory requirements," says Triphonas Kyriakis, head of analytics for Europe, the Middle East and Africa at MSCI.