Lorenzo Naranjo offers a case study of the Chilean corporate bond market to show how difficult pricing can be in illiquid markets, and tenders a solution

Liquidity is a major issue when investing in corporate bonds, a problem that gets amplified in the context of emerging markets. Liquidity by itself is an elusive concept that means different things for different people. In a nutshell, liquidity could be defined as the ease of trading a security. Hence, there are several dimensions that affect the liquidity of a particular bond - transaction costs, price impact, and search costs, among others.

Direct transaction costs are usually measured by bid-ask spreads, and are typical of dealers' markets in which one or several dealers quote bid-ask prices for a set of bonds. This is usually the case for on-the-run bonds. More liquid bonds will then exhibit lower quoted bid-ask spreads.

Price impact relates to the depth of the market in which a security trades. In a deep market, large quantities of bonds can be purchased or sold without affecting the price.
Finally, in many over-the-counter (OTC) bond markets, it might be difficult to find a counterparty or dealer that is willing to buy or sell a specific security, in which case the liquidity of a particular bond issue is going to be low.

The liquidity of a bond might affect the price and the frequency of transactions. For example, if investors think that it will be more expensive or hard to sell a bond in the future, they will com-mand a lower price in order to hold the bond in their portfolios. To obtain an immediate execution of a sell order, investors will have either to sell at a discount, or wait until a suitable buyer wants to buy the bond.

There is now a well-established body of empirical literature that shows that liquidity is an important determinant of corporate bond yield spreads and returns. In many cases, the liquidity component of corporate bonds can explain a larger fraction of the yield spread than the default component itself. Even in the US, a 2008 study by Sriketan Mahanti and colleagues found that the median corporate bond trades every 12-18 days, out of 22 trading days per month. These figures contrast sharply with the ones obtained for equity markets, in which even less liquid stocks trade many times a day. Furthermore, the authors report that in their sample 10% of the outstanding corporate bonds trade, at best, two to three times a year.

Emerging markets
In emerging markets the liquidity of corporate bonds is even lower. This poses several problems for investors wishing to profit from apparent higher yields in these markets.
There are additional sources of risk that affect corporate bonds in emerging markets. The first one is the so-called liquidity risk. In times of market stress, the ability to sell quickly might be hampered, since these bonds do not trade frequently enough even in normal times. Unwinding large positions might become more expensive during crises than in normal times.

A second source of risk for foreign investors is the possibility that the bonds are issued in a currency that is different from their own. In times of market stress, the correlation between the liquidity and the strength of the currency in which the bonds are denominated might be positive, compounding the risk for foreign investors.

Third, inflation in emerging markets might be higher than in developed markets, inducing an additional source of uncertainty that might be negatively correlated with the liquidity of the bonds.

Besides the price risks taken on by investing in such securities, the liquidity of corporate bonds in emerging markets also generates additional problems for the back office. Since most bonds do not trade every day, marking positions to market might be quite challenging. This is an important issue for open-ended funds from which clients might withdraw money at any moment. The same is true for computing risk measures.

Since in many emerging markets not even government bonds are liquid enough to trade every day, the daily pricing of corporate bonds is an important obstacle - and an important issue to keep in mind - when investing.

An example of an emerging corporate bond market is the case of Chile. It is fairly liquid when compared with other emerging markets, given that there is a large and active pension fund industry and several other open-ended mutual funds. However, the liquidity of the bond market is not high enough to obtain daily prices in order to mark-to-market corporate bond portfolios.

There are several solutions. First, following the methodology used to compute value-at-risk (VaR), one could first group bonds by credit ratings and try to compute the term-structure of corporate bond yields for each credit category. Figure 1 shows the observed prices for AAA bonds with an approximate duration of 10 years from 25 June to 26 July 2012. The data is taken from RiskAmerica.com, a startup firm owned by the Pontificia Universidad Catolica de Chile that has specialised in solving these issues.

As can be observed from the figure, grouping bond prices by credit ratings does not solve the problem completely since, for many days, there are no observed prices for this particular maturity. This is actually the same problem that exists for government bonds in the Chilean market. There are not observed prices for all maturities every day.

In order to estimate the term structure of government bonds using all the price information that is observed every day, in 2006 Gonazlo Cortazar, Eduardo Schwartz and I proposed a methodology that takes advantage of the time-series dynamics of the whole-term structure of interest rates to obtain daily estimates of bond prices and yields.

Using the same insight, it is natural, then, to extend the methodology to estimate the term structure of corporate bonds for different rating categories, using all the observed prices on a given day. Ideally, one might want to use all prices for government and corporate bonds together to take advantage of the possible correlations among them when estimating the different term structures.

Figure 2 shows the time-series from 25 June to 26 July 2012 of the estimated yields for AAA, AA and A bonds with duration of 10 years. It is interesting to note that there is a marked difference between the yield spreads of the three rating categories that varies over time, but the methodology seems reasonable and useful for practitioners investing in this market.

As we have seen in this article, investing in corporate bonds in emerging markets requires solving some complex issues related to the liquidity of such instruments.
First, it is important to understand the liquidity risk involved in buying emerging market corporate bonds. We understand better today the effects of liquidity and liquidity risk on asset prices and, as such, we should take into consideration the implications of such effects for illiquid assets in thin markets.

Second, the pricing of such assets is difficult and requires some care. Recent advances in financial theory allow us today to use available price information efficiently and overcome, at least partially, the low frequency of transactions.

Finally, a clear understanding of the institutional details of each market is crucial for a serious implementation of an investing strategy in such markets. It is important to know the players, the size and volume of the market, and the sources of information available to investors.

Lorenzo Naranjo is a professor in the finance department of ESSEC Business School