COMMENTS ON OPTIMAL SHARING CONTRACTS

DISCUSSANT: NADEEM UL HAQUE
SPECIFIC COMMENTS

I was indeed very pleased to see the sophisticated modeling techniques that Mr. Al-Suwailem is using in his paper. Indeed he has quite comprehensively covered the issue of Islamic investment sharing contracts. I would urge him to review our work in the KM book and relate his paper and work to it after all academic work must seek to build on previous efforts. But more on that later. Let me first review the paper and its technical aspects.
In my comment, I will first review the paper and its technical elements and then make a more general comment on this line of research as well as research in Islamic economics and economies.
The author has built a nice model to study the investment process and conduct a sort of “horse race” between the neo-classical and Islamic approaches. He is even able to consider various alternative mechanisms of finance examine the underlying assumptions and draw some implications for converging to equilibrium. There is therefore a lot here for us to develop and debate more fully the implications of this model.
In my view, the author has presented us with three papers instead of one only to make my task more difficult to the glee of the organizers who are always looking to maximize the conference participant’s effort. In my comments, I am indeed going to suggest that this long version be split up into three papers so that a clearer focus can be achieved in each of them. With that general comment let me look at the specifics of the paper.
In looking at the basic model, the most important point to note is that the effort enters the model in an asymmetric manner: it costs the firm more in bad states and less in the good state. There is a direct cost of effort measured by c(e) which is what it costs the firm for inducing a better effort from the employee/manager and this direct cost is not state-dependant. In the good state this is the only cost. However, in the bad state, the firm faces and additional cost, Rl(e). This modeling artifact is driving a fair number of results and must be clearly motivated. It is this asymmetric cost of effort that leads to sub-optimal behavior even in the case of self financing and (equation 3 and property 1). If Rl(e) were did not depend on e, we would have the first best solution of self financing as λ(ef)would be equal to 1. It is intuitive to think that the cost of effort should not depend on the state of nature.
However, I do concede it is a modeler’s prerogative to decide how to capture reality. If it is a case of attempting to capture inefficiencies arising from uncertainty, we know that risk aversion alone will do it. We do not really have to assume that when nature deals a bad hand, effort will compound the impact. If this is the case then surely one can think of introducing some form of incentive compatibility constraint to induce a first best situation which could easily be done if we constrain the optimization problem of (1) by the constraint that Rl(e)=c where c is some arbitrary level or an extremum of Rl(e). In this case λ(ef) =1 and the first best obtains.
In section 3, sharing with symmetric information, and in section 4, Debt contract with symmetric information, it is shown that the sharing Pareto dominates debt when the expected cost of bankruptcy is significant (ie., h ≤ (1-q)b) (property 8). But the way it is modeled, bankruptcy only visits a debt contract and not a sharing contract. It seems to me that for a valid comparison the cost of bankruptcy have to be symmetrically applied to both sharing and debt. For sharing, monitoring costs could be equivalent to bankruptcy costs and hence the reason for introducing symmetry into the cost structure. It is not surprising then that Table 1 shows that the entire difference between the 2 forms of finance is owing to the asymmetric treatment of bankruptcy.
In section 5, Asymmetric Information, I expected to find contracting approaches that would be based on market monitor-able indicators such as the share price or the value of output. Instead the paper chose the audit approach. Auditing is costly and time consuming, whereas contracting approaches align the incentives of the parties concerned and produce outcomes that are beneficial to both.
Once again we see that the bankruptcy costs are an important reason for the differences between the sharing and the debt problems. In addition, the paper makes several observations that seem to fly in the face of conventional wisdom. For example, “lender is more likely to audit than the financier” (page 24) and that there is a “lower cost of auditing under sharing.” The debt contract is preferable to sharing in conventional economics because it frees the lender form the need to monitor and audit because of the expectation of a guaranteed return. In other words the lender only faces a default risk and not an agency problem. The debt contract has to be honored regardless of the state of nature whereas the sharing contract is state dependent. Moreover, debt has seniority over sharing and hence provides the lender with more security. To visit bankruptcy costs only on the debt contract in a sense makes the shareholders senior to debt holders.
In the next section, Extensions several interesting new elements are introduced and serve to illustrate to us the sensitivity of the model to changes in the underlying design. First the probability of success is treated as an endogenous variable by making it an increasing function of effort. Now effort is working in the right direction, not only to increase output but also to increase the success of realizing that output. This is perhaps the most realistic and interesting case to examine the agency problem and its effect on the two alternative forms of financing. But the incentive compatible sharing contract has to be contrasted with the debt contract when effort is unobservable to the financier. The solution that is derived suggests that there are “two sources of inefficiency of debt contract: sub-optimal level of effort in addition to bankruptcy costs” (page 31). However as Figure 3 shows the debt contract induces more effort but with a lower profit because bankruptcy costs are only associated with it.
The second extension is endogenising bankruptcy cost. This is done by making bankruptcy and increasing function of effort. Now effort appears to be running in two different directions. In the one case it is working as expected in increasing output and returns to the project but also it increasing the cost of bankruptcy. This is the Drexel Burnham Lambert case, where the managers had put in place the poison pill so that in case of bankruptcy, shareholder and the lender both lose out to management. But since bankruptcy costs only affect the debt contract the solution is obvious that this new arrangement will make debt less attractive.
Having prepared the ground with an understanding of the pure debt and sharing contracts, the paper examines the Musharakah and Mudarabah cases in section 8. These should perhaps form a shorter paper building off another paper that we have already discussed above. Here the interest would lie only in examining the various sharing ratios and perhaps deriving and inventory of the conditions under which the mudarabah would be better than the Musharakah. For example, we could think of states of nature and the specificity of effort (professional expertise) as a form of capital that could lead to a partnership or pure financing.
In the final section, Evolution of Sharing the paper examines the financing relationship as a repeated prisoner’s dilemma and comes up with the familiar solutions of ‘tit for tat’ and how repeated interaction will develop cooperation and reputation. The manner in which the problem has been set into a prisoner’s dilemma setting is quite imaginative but unfortunately the analysis does not appear to take us beyond the well know prisoner’s dilemma results. As I said, this is the third paper but perhaps it the one that will require a little more effort and further conceptualization. If the purpose is to induce incentive compatible contracts from repeated interaction, perhaps the standard contracting models are the better way to go. Even in game theory rather than prisoner’s dilemma, I would consider mixed strategy and signaling games. This paper perhaps will require some further work to clarify the objective as well as the modeling strategy.
GENERAL COMMENTS
These are my technical remarks on the paper or should I say three papers. I do hope that this will benefit the author and allow three nice publishable papers to emerge. Let me now make one general comment on Islamic economics as someone who has been associated (though not deeply involved) with it over the years.
For many years now, Islamic economics has been preoccupied with proving the viability of the interest-free system. Several papers of a fairly sophisticated nature were prepared in the eighties –several of them at the Fund—to understand the Islamic Interest free banking system. To establish the subject of Islamic economics, many papers attempted a comparison of the neoclassical or the conventional model with the Islamic interest-free model (the horse race approach). The stability properties of the Islamic economy and the impact of increased uncertainty on investment and savings were of particular interest.
A very important collection of these papers (Reference 1 below) was published in the late eighties and should by no be standard reading in this area. I would urge Mr. Al-Suwailem to look at that volume and relate his work to some of the papers there. In that volume, Mohsin Khan’s paper showed that the Islamic banking system under certain conditions might even show more stability than the conventional economy. The question then was what sort of supporting infrastructure such as central banking system and fiscal financing instruments could be made available.
In that same volume Abbas Mirakhor and I published 2 papers that relate very closely to the paper that we are discussing now. In Optimal Profit-Share Contracts and Investments in an Interest-Free Islamic Economy, we argued that under conditions of uncertainty and asymmetric information, a fixed interest rate performs a function of protecting the investor and hence may lead to superior outcomes than the system without fixed interest rates. However, when one allows for contracting, the interests of the investor and the entrepreneur are aligned so that they operate like one residual income earner. The potential conflict between borrower and lender is eliminated. Investment levels may therefore be higher under a system of equity sharing with no fixed interest. We did end the paper by noting that to make this system work, we would need “extensive legal and institutional infrastructure that will allow smoother transactions with enhanced monitoring and speedy enforcement,” and that we would need the development of financial markets so that “secondary markets develop fro the trading of profit-sharing contracts.”
In our other paper Saving Behavior in an Economy Without Fixed Interest we suggested conditions under which the assertion that “the elimination of interest will reduce savings” may not be true. Once again, we concluded that “a stronger conclusion may obtain if the effects of structural changes are put in place.”
What has surprised me over the years is that considerable interest continues to be taken in the early questions and a somewhat disproportionately lesser effort is put into the issue of supporting structural changes that Abbas and I pointed to many years ago. To me the greater payoff is in designing the supporting infrastructure for a well functioning Islamic financial market. The recent financial crises have amply illustrated this point.
Islamic forms of finance are by now fairly well established. Major banks like Citibank have established subsidiaries for Islamic modes of finance. Morgan Stanley has an index for measuring the performance of Islamic instruments. Abbas Mirakhor and I designed an Islamic instrument for government finance that Iran has adopted.
In view of this trend of acceptance of Islamic instruments, the need for ‘race horse’ models (comparison of debt and sharing) is now somewhat limited. Instead, I would argue more effort should be put into developing understanding and developing financial markets for economic growth in Islamic countries. In my view Islamic economics should not be merely economics of interest-free banking but also the economics that will help Islamic countries. Moreover, whether we do conventional or Islamic banking there is a lot of common infrastructure (legal, monitoring technology, and market development) that we need to review, research and understand. We must make more of an effort in that direction for the prosperity and survival of our people.
REFERENCE

Khan, M.S. and A. Mirakhor (eds)(1987), ‘Saving Behavior in an Economy Without Fixed Interest’ and ‘Optimal Profit-Share Contracts and Investments in an Interest-Free Islamic Economy’, in Theoretical Studies in Islamic Banking and Finance, Houston: Iris Books.