Charta is on Google Play: practise on the go. Get the app
Charta

Account for the potential difficulties fund managers meet in managing liquidity risk in fixed-income portfolios, data modeling included.

FRM® Part II · Risk Management and Investment Management · Liquidity Risk Management · FRM2.IM.16.b

The idea

Fixed-income funds face harder liquidity management than equity funds because bonds trade over the counter through dealers, with no central order book, and most individual issues trade rarely. Many holdings are valued with evaluated or matrix prices taken from comparable bonds, so the reported NAV can overstate what a sale would realise. Those smoothed prices also understate volatility and correlation when used for risk estimates. Observed trade data are biased toward small trades in calm, liquid conditions, so average costs understate the cost of selling large positions in stress. Dealer capacity also shrinks when everyone sells at once, and cost per unit rises with size. A practical measure of the time needed to sell a position is Days = position ÷ (participation rate × expected daily volume). The usual error is assuming normal-times cost holds in stress.

Formulas

Days needed to liquidate a position

Days = position ÷ (participation rate × expected daily volume)

position
the size of the holding
participation rate
the maximum share of daily volume the fund will trade
expected daily volume
the expected volume per day, counting days with no trades

For rarely traded bonds use the expected volume per calendar day, the share of days traded times the volume when it trades. Average volume on trading days overstates capacity.

A practice question

From the question bank, with the reason each wrong option is wrong.

Why is managing liquidity in a corporate bond portfolio generally harder than in a portfolio of large-capitalisation equities?

  1. Bond prices are quoted as yields, which makes it impossible to place an executable orderWhy not: Quote convention does not determine liquidity; the difficulty is market structure and thin trading.
  2. Bond investors are prohibited from using dealers and must find a counterparty themselvesWhy not: Reverses the market structure: dealers are the main intermediaries in bond markets.
  3. Over-the-counter dealer trading, no central order book, and many issues that rarely trade - correct
  4. Credit risk cannot be priced until the bond matures, so bonds cannot be traded at fair valueWhy not: Credit risk is priced continuously in spreads; it does not stop trading until maturity.

Corporate bonds are dealer-intermediated, fragmented over thousands of issues and often trade only occasionally. That structure, not quote conventions or credit pricing, drives the difficulty of liquidity management.

A harder one

A risk team estimates the trading cost of its corporate bond holdings using only recorded trades. Many holdings trade rarely, and the recorded trades are mostly small and executed on calm days. What is the main bias in the resulting cost estimate?

  1. It understates stress-time cost of large sales, because the sample favours small, calm-market trades - correct
  2. It overstates cost because rarely traded bonds only appear in the data when dealers quote very wide spreadsWhy not: Assumes the sample is skewed toward illiquid conditions, whereas it is skewed toward small, calm-market trades.
  3. There is no bias because the average of observed costs is an unbiased estimate for the whole portfolioWhy not: Ignores that the sample is selected, not random, and that cost rises with trade size and in stress.
  4. It biases volatility estimates but has no effect on cost estimatesWhy not: Confuses stale return data with transaction cost data; selection in trade data biases cost directly.

Observed trades are a selected sample: small, in calm periods, in bonds that do trade. Applying their average cost to large positions in stress understates what liquidation will cost.

The hardest tier

A fund holds USD 12 million of a corporate bond. Over the past year the bond traded on 40% of days, and on days it traded the average volume was USD 5 million. Policy limits the fund to 25% of daily volume. Using the expected volume per calendar day, the number of days needed to liquidate the position is closest to:

  1. 24 days - correct
  2. 9.6 daysWhy not: Uses the volume on trading days (5 million) and ignores that the bond trades on only 40% of days.
  3. 6.0 daysWhy not: Ignores the 25% participation limit and divides by the expected volume per day of 2 million.
  4. 60 daysWhy not: Applies the 40% trading frequency twice, giving an expected volume per day of 0.8 million instead of 2 million.

Expected volume per calendar day = 0.40 x 5 = 2.0 million. The fund may take 25% of it, 0.5 million per day. Days = 12 / 0.5 = 24. Rare trading and participation limits both lengthen liquidation.

In the app: 6 practice questions on this outcome, spaced review that brings it back before you forget it.

Start free →

More in Liquidity Risk Management