The traditional problem of Liquidity Risk is that the data needed for calibrating these models is only available for liquid instruments, trading on a regular basis and for which books of bid/ask and volumes are available. For this reason the current approaches to measuring Liquidity Risk fail providing any indication for the most opaque and illiquid instruments, or where the measurement of Liquidity Risk is mostly needed.

The new approach introduced here is based on liquidity scenarios, which is universal, because it covers potentially any financial asset, from equities, to bonds, to OTC derivatives under a homogeneous and consistent approach. Each scenario simulates a defined set of liquidity conditions applied transversally to all assets. Three liquidity scenarios are available: Normal Market Conditions, Stressed Conditions, Highly Stressed Conditions.

In every scenario a number of parameters can be controlled, influencing the five components of liquidity risk identified in this paper:

• Fair Value Bid / Fair Value Ask;

• Pricer Haircuts;

• Outstanding Notional;

• Equity Market Capitalisation;

• Percentage of Stock Ownership.

Fair Value Bid / Fair Value Ask. When an instrument is illiquid how does a market-maker create a price for it? This question has inspired and driven our research on liquidity risk. Professional market-makers rely on pricing functions to derive the fair value of a financial instrument that is not actively traded on a market.

When they need to quote a bid or ask for that instrument, they insert in their pricing functions the bid/ask of the underlying derivative market data.

Let’s make an example for a convertible bond. The pricing function for such instrument will require as inputs:

• Interest rate swaps to derive the Libor discount factor curve;

• Credit Default Swaps (CDS) to measure the credit risk of the bond guarantor;

• Underlying stock price;

• The implied volatility of the embedded call option.

In building the bid for the bond, the market maker will choose between the bid and ask of the above data inputs (or risk factors), depending on what action he needs to take to hedge that risk. For example in buying the bond he will acquire a short position on the credit spread of the bond, facing the risk of an issuer default. If he wants to hedge that risk he will buy protection on the relevant CDS, hitting the spread Ask in the market. Therefore, when inserting the CDS value to quote the convertible bond bid, he will insert the CDS spread Ask.

Similarly, he will insert the implied volatility bid in the function, as a long position in the bond coincides with a long position on the option volatility. To hedge that exposure the market-maker will hit the bid of that risk factor.

Our liquidity scenario replicates this process for every bond, derivative, option, certificate and in general for any instrument that holds a pricing function. The liquidity risk calculator understands what is the exposure of an instrument to each of the risk factors and uses the bid or the ask of the risk factor accordingly.

Pricing Function Haircut. Liquidity is normally an inverse function of complexity: the higher the complexity of the underlying financial instrument, the lower the liquidity. This component allows the user to penalise certain asset classes, identified by a Pricer tag. I.e. ABSs are identified by a dedicated tag and the liquidity scenarios increase their liquidity risk by a specific haircut.

Outstanding Nominal Haircut. For bonds and certificates the amount issued (outstanding nominal) can influence the underlying liquidity. The bigger the issued notional, the greater the associated liquidity will be. The liquidity scenarios offer the possibility of configuring a grid of liquidity haircuts by outstanding nominal bucket, divided by currency.

Market Capitalisation Haircut. Liquidity in equities is a function of market capitalisation, and typically the greater the market capitalisation, the greater the liquidity. In the scenarios, users can define a grid of liquidity haircuts by market cap buckets.

Equity Ownership Fraction Haircut. Another source of liquidity risk in equities is the amount of the company that the portfolio owns. E.g. a portfolio that owns 10% of a company faces higher liquidity risk than one that owns 0.01% of the same stock. The liquidity scenario contains a grid of liquidity haircuts associated to different levels of market capitalisation ownership.