Preventing Losses Associated with Asymmetric Information in Loans

LoanOS resolves many of the information asymmetries that pervade the syndicated loans industry.

Asymmetric information or “information failure” can severely hinder the efficient operation of credit markets. This occurs when one party involved in a transaction has more knowledge than the other party, tending to result in a variety of inefficiencies and losses.

The leveraged loan market is particularly opaque and is rife with examples of asymmetric information. Trades in this market occur over the counter without centralized pricing, and some loan characteristics are unobservable until a market participant is actually in possession of a loan. As a result, the purchaser of a loan tends to have significantly less knowledge about the loan characteristics than the seller.

Forest Park’s LoanOS offering has the potential to mitigate information asymmetries such as these. LoanOS accomplishes this by providing a centralized platform on which information about trades’ and loans’ characteristics is more publicly available than it is under the status quo.

There are two channels through which asymmetric information can impact loan pricing and the allocation of credit. Thus, LoanOS acts to improve issues in these two areas:

First, asymmetric information is currently manifested directly in the primary market for loans.

In other words, it is reflected at loan origination. More highly-informed buyers in the primary market extract rents from less-informed ones, depressing the price that less-informed market participants are willing to pay for a given loan.

In order to quantify the effects of asymmetric information in this model on pricing, we need to know the following:

1. The probability of default on a loan
2. The expected loss given default
3. The probability that information is asymmetric

Second, asymmetric information also impacts loan pricing in the secondary market.

Occasionally, market participants will become financially distressed or will require liquid funds for other reasons. In this case, they will want to sell their assets. However, market participants may also want to sell assets because they receive unfavorable information about said assets. This implies that sellers of assets will have to sell at a discount that increases in the severity of the information asymmetry.

It is also possible to quantify the losses from this form of adverse selection by estimating the cost of distress as well as other fundamental parameters (such as the average settlement time and discount rate). Importantly, this model can be used to say something about how a reduction in settlement time reduces the adverse selection discount.

Present economic literature supports that asymmetric information negatively impacts the primary and secondary loan markets. For instance, a paper written out of Harvard Business School, Asymmetric information effects on loan spreads (Ivanshina 2008), estimates that the cost of information asymmetries in loan syndications can be around 29 basis points (bps). This was calculated from a measured 9% change in lead share (the percentage of the loan retained by the lead arranger at the loan origination). This is equivalent to a $24 million increase in the lead bank’s exposure, an amount which can be of economic significance for a bank.

Thus, Forest Park’s LoanOS software is posed to save its users significant sums of money by reducing information asymmetries. Specific savings as a result from the improved symmetry of information associated with LoanOS are computed with the following calculation, which states that the deadweight loss associated with a particular trading event is:

In this equation, there is an asset that pays a dividend vdt each instant, and a continuum of investors who discount at a rate r. At any time, investors may transition between patience or impatience with Poisson intensity λ. Patient investors receive the full value of the dividend, and impatient investors are paid only a fraction Θ of the dividend when holding the asset. A buyer learns the quality of an asset according to a Poisson process with arrival rate (1/τS), where τS is the time required for a buyer to distinguish a good asset from a bad one.

This can be simplified so that the deadweight loss per trade with separation time (τS) is:

Then, deadweight loss over the life of an asset can be calculated. If the asset matures at the rate (1/τM) and trading events occur at a rate λ, then the expected number of trading events is (τM/λ). Thus, the total deadweight loss is:

Given typical values of τM = 7 years, λ = 2 trades per year, 1 — Θ is a discount of 10%, r = 0.02, and τS = 30/360 years (i.e. a settlement time of 30 days), then the total loss is:

Thus, we calculate the losses associated with asymmetric information in this scenario to be about 300bps.

By using our LoanOS software to reduce the τS to (7/36), the loss is instead computed as 68bps. This is a substantial savings: over 200bps, or two thirds.

For more information about Forest Park and LoanOS, please refer to our website at forestparkgroup.com or reach out to us via email at opportunities@forestparkgroup.com.