The Measurement Research On Risk Of Security Investment Based On Bayes Bilateral Integrated Moment

In order to measure risk practically, in 1952 Markowitz put forward using the variance of return rate as risk measure index based on the definition of risk which is the uncertainty or fluctuation. The variance risk measure indicates the threshold of risk measure, and it's adopted proverbially because its familiarity and facility in mathematical manipulation. But many scholars criticized variance risk measure because the strict hypothesis of M-V approach and the symmetrical nature of variance which assigns the same weight to positive as to negative deviations from the expected value does not capture the common notion of risk as a negative undesired characteristic of an alternative.The downside risk approach which consider only the loss proportion of return could be took into account in measuring risk vibrate M-V approach in deed. Therefore, the downside risk catch investors' mind better, and it seems to be more appropriate for measuring risk. But the downside risk which solely considers underperformance as risk and discard the better investment opportunities, and the partial nature of downside risk which assigns neutrality to returns above target rate does not capture the pursuits of better investment opportunities by most investors.Considering investors' different attitude on positive deviations and negative deviations, a risk measure index should catch investors' mind accurately and reflect essence of risk. On the other hand, the probability of positive and negative deviations are derived from data usually, moreover, uncertainty from inadequacy of information will increase the difficulty on making investment decision. So this dissertation designs a new risk measurement index: Bayes Bilateral Integrated Moment(BBIM). Unlike variance risk measure which punishes the desirable upside movement as hard as undesirable downside movement, unlike the downside risk which solely considers underperformance as risk and discard the better investment opportunities, BBIM are intuitive measures of risk that not only focus on return dispersions below a specified target or benchmark return, but also use the return dispersions above a pre-specified target late that contains promising profit. Meanwhile, through Bayes method we can absorb new information, revise the return distribution from transcendental distribution to posterior distribution, so we can get a more accurate and integrated distribution of the return rate of the securities, and then using it to make a good investment decision. What's more, BBIM are a class of full domain risk measures, where downside risk issupplemented with the "upside potential" and it uses Bayes theory to reduceuncertainty.This dissertation is organized as follows: chapter 1 is the preface, the motivation, path, content and innovation are presented in this chapter. Chapter 2 discusses the nature attributes of the investment risk from the definition of risk we already have now, and then we research the nature definition and the character of the security investment risk. Chapter 3 is a literature overview of the risk measure, we emphasis on the traditional mean variance approach and the downside risk approach. Chapter 4 is the central part of this dissertation, where the bayes bilateral integrated moment is established on the base of Bayes theory, downside risk and upside potential. According to the variation coefficient which is based on the M-V approach, we substitute lower partial moment(LPM) for the numerator of variation coefficient, and use higher partial moment(HPM) as the substitute of the denominator, then we got bilateral integrated moment. Subsequently, through Bayes method we can absorb new information, revise the return distribution from transcendental distribution to posterior distribution, so we got bayes bilateral integrated moment. Meanwhile the theoretic evidence is given in this chapter. And the demonstration is found in chapter 5.