Asset Pricing on Segmented Markets: A Synthesis, an Extension and an Application to Islamic Financial Markets
The starting point for this thesis was that although Islamic financial assets and Islamic financial intermediaries have grown into relevant players, they are highly dependent on one another since funding of financial intermediaries has typically been achieved through Islamic investment accounts, which are profit-sharing-based contracts and represent 67% of Islamic banks’ funding. This reliance on profit-sharing-based contracts—that do not guarantee fixed interest payments—comes at a cost, namely, the risk that inadequate rates of return could lead to massive withdrawals that may reach systemic proportions and cause concern on the part of supervisory authorities as expressed in IFSB Guidance Note 3, Article 9.
Consequently, the main objective of this thesis was to develop an asset pricing (valuation) formula for Islamic financial assets that captures the segmented market nature of financial markets where Islamic banks operate and solves their adequate returns benchmark problem.
In the second chapter, we analyzed the cash flows and risks involved in Islamic financial contracts and found in the case of Islamic investment accounts and Sukuk that their cash flows and risks depend on a two-stage structure. On the one hand, their cash flows and risks hinge on their underlying contracts (first stage). Mark-up contracts are able to secure riskless cash flows while profit-sharing contracts are unable to do so. On the other hand, these cash flows are then subject to a number of transformations (second stage) such as smoothing, management fees, reserve creation, and pooling of different investments. These transformations may alter the stochasticity of the cash flows distributed to depositors/Sukuk holders in a sense that individually riskless contracts become slightly risky and individually risky contracts become slightly less risky.
In the third chapter, we obtained four main results: First, we successfully derived valuation formulas for all assets available (including Islamic financial assets) on different levels of market segmentation. The required expected return on common assets (Islamic financial assets and Islamic stocks) is computed in an identical way as the classical CAPM with the exception that rather than taking a single riskless rate as the return on riskless assets, a mixture (weighted by the aggregated risk preference parameters of both investor groups) of the riskless rates available to Conventional investors and Islamic investors (assumed to be an interest rate of zero in our model) should be used. The required expected return on the restricted asset class (non-Islamic stocks) consists of a single riskless rate, namely that of the unrestricted group (Conventional investors) plus a risk correction that is based on the risk preferences of the unrestricted group and an additional term that reflects the demand frictions caused by the fact that Islamic investors cannot invest in non-Islamic stocks (a demand-effect term). Second, valuation formulas that contain unobservable quantities (risk preference parameters) and an explicit reference to the riskless rate cannot be used to value Islamic financial assets in practice. Hence, we express the valuation formulas only in market-observable quantities independent of the riskless rate. Using the reformulated valuation formulas, we can show that the valuation for common assets is no longer a linear function of the expected return of the market portfolio as is the case in the classical CAPM; instead, we observe a linear two-factor valuation model for Islamic assets and Islamic stocks. For the restricted assets (non-Islamic stocks), the linear market portfolio structure breaks down completely resulting in a non-linear two-factor model of valuation. Third, statistical significance analysis found that the security market lines of the valuation formulas that overlook the segmented market framework are never identical to those of the theoretically correct valuation formula. For the valuation of specific assets, however, there are exceptions: Assets whose covariance/risk lies exactly at the intersection point of the security market lines for segmented (correct) and unsegmented (incorrect) markets have the same required expected return. We conveniently call this effect of an accidentally correct valuation result even if a wrong valuation formula is used the “double error compensation effect”. Fourth, we test the economic significance, i.e., whether valuation errors from using an incorrect valuation model are large enough to induce economic consequences. For this analysis we use a sample of representative sub-market portfolios as examples of specific assets. We find that the differences in the required expected returns between the theoretically exact segmented model and the unsegmented market model are nearly always economically significant when transaction costs are used as benchmarks. With other benchmarks mixed results are obtained.
Finally, in the fourth chapter, we obtained two results. Our first result is that we determine over-, correct, and undervaluation for both short- and long-term using full-sample and a five-year rolling estimation window for 81 Islamic banks in 16 countries. Based on these valuations—second results—we develop recommendation for practical application. On the one hand, a traffic light system for private investors is developed that translates valuation result into withdraw, hold, and deposit funds recommendations. On the other hand, for institutional investors no standardized system like a traffic light system is needed in general because institutional investors are assumed to possess a high degree of financial literacy. Therefore, only the necessary input data required for computations are provided. Only if the regulator is concerned about systemic risk of the Islamic financial system, the regulator might wish to assure that Islamic banks do not invest in overvalued Islamic investment accounts. In that case, a traffic light system might come into play. Finally, since transparency is connected with the reliability of the Islamic financial system, the traffic light system must be reliable as well. Consequently, regulators or central banks publishing the traffic-lights-system should do so periodically and include the valuation on one (web)page because only then comparisons of different banks’ Islamic investment accounts will become possible.
Our results have a number of practical implications. The first one relates to empirical research in connection with asset pricing models in general and segmented markets in particular. Our asset pricing formulas show that valuation formulas on segmented markets that consist of observable quantities only comprise at least two market factors. Therefore, required expected returns cannot be determined using regressions that contain just one market factor (even when combined with Fama/French and Carhart factors); instead, at least a second market factor must be integrated into the analysis. Even then, the factor loadings of the segmented markets asset pricing models are not identical to regression coefficients in general. Only if a specific model of asset returns is used, namely asset returns that are a linear function of the return of the market portfolio and another factor that is uncorrelated with the market portfolio’s return, will regression coefficients result (see Errunza/Losq (1985) for such a model). Second, our pricing formulas contain a valuation model for Islamic financial assets that does not contain any reference to the riskless interest rate r, thus, are indeed Shariah-compliant. In other words, Equation (7b) is the asset pricing formula for Islamic financial assets that has been missing in the literature so far. In particular, it offers an alternative valuation of Islamic financial assets that does not rely on mimicking Conventional rates. Hence, it takes into account the criticism of the recent AAIOFI Standard 27 on Indices, Clause 7 as well as decision number 76 (7) of the 8th conference of the International Islamic Fiqh Academy of Saudi Arabia, which took place in Brunei 1993 that Conventional interest rates should not be used as a benchmark for Islamic assets (International Islamic Fiqh Academy, 1993; AAOIFI, 2010: 489). We highlight the notion that a country-wide Islamic returns benchmark is not a very reliable index for valuation since it does not take into consideration the unique risk profile of each individual Islamic asset i, which would result in a unique required expected return for each Islamic financial asset. Third, we provide financial institutions with an alternative to tweaking the returns of profit-sharing products. Consequently, there remains no need to smoothen the returns and mimic those of Conventional deposits and bonds which was done in order to remain competitive or to avoid mass withdrawals by depositors. These return transformation techniques came at a high cost: (i) mimicking the returns of conventional deposits violates the spirit of Shariah conformity (International Islamic Fiqh Academy, 1993; AAOIFI, 2010: 489) and is therefore not sustainable in the long-term; (ii) it exerts additional pressure on the bank by forcing them to meet the returns of Conventional deposits if the returns on actual investments were not high enough; (iii) they give rise to displaced commercial risk which is “the risk arising from assets managed on behalf of IAH (investment account holders) which is effectively transferred to the (bank’s) own capital because the (bank) follows the practice of (smoothing) when it considers this necessary as a result of commercial and/or supervisory pressure” (IFSB GN-3: 3); (iv) smoothing practices have inherent inter-generational reserve problems: Reserves that have been built up in the past and are used today for the benefit of the current investment account holders, who may be different than those who contributed to the reserves in the past; (v) smoothing conceals the actual returns achieved by bank management and removes the ability of regulators and depositors to evaluate the quality of investment management at the bank, (vi) smoothing only hides the problem of fluctuations in the returns of investment accounts from the depositors’ perspective, yet the banks must deal with these fluctuations and must determine the correct amount of smoothing and return transformation to apply. By removing the need for smoothing, transparency regarding Islamic investment accounts can be guaranteed and might help supervisors to monitor the stability of the Islamic financial system. On the one hand, Islamic banks who offer overvalued Islamic investment accounts might be confronted with withdrawal risk in the future or, at least, will have problems getting enough funding in the future. On the other hand, Islamic banks are closely connected since they invest funds in Islamic investment accounts of other banks. A repeated investment in overvalued Islamic investment accounts by some banks might indicate a potentially dangerous investment chain.
Philipps-Universität Marburg
2018-03-29
Marburg
Universitätsbibliothek Marburg
UB Marburg
http://archiv.ub.uni-marburg.de/diss/z2018/0103
https://doi.org/10.17192/z2018.0103
urn:nbn:de:hebis:04-z2018-01034
Asset Pricing on Segmented Markets: A Synthesis, an Extension and an Application to Islamic Financial Markets
en
Badreldin, Ahmed
Universitätsbibliothek Marburg
mailto:auskunft@ub.uni-marburg.de
https://archiv.ub.uni-marburg.de/adm/img/unilogo-dfg.gif
https://www.uni-marburg.de/bis
https://archiv.ub.uni-marburg.de/diss/z2018/0103