2013数学中国国际赛论文参考模版 3

第二届“认证杯”数学中国

数学建模国际赛 承 诺 书

我们仔细阅读了第二届“认证杯”数学中国数学建模国际赛的竞赛规则。 我们完全明白，在竞赛开始后参赛队员不能以任何方式（包括电话、电子邮件、网上咨询等）与队外的任何人（包括指导教师）研究、讨论与赛题有关的问题。

我们知道，抄袭别人的成果是违反竞赛规则的, 如果引用别人的成果或其他公开的资料（包括网上查到的资料），必须按照规定的参考文献的表述方式在正文引用处和参考文献中明确列出。

我们郑重承诺，严格遵守竞赛规则，以保证竞赛的公正、公平性。如有违反竞赛规则的行为，我们将受到严肃处理。

我们允许数学中国网站(www.madio.net)公布论文，以供网友之间学习交流，数学中国网站以非商业目的的论文交流不需要提前取得我们的同意。

我们的参赛队号为：1273 我们选择的题目是： A 参赛队员 (签名) ：

队员2：金乔

队员3：刘家栋

参赛队教练员 (签名)： 刘保东

队员1：朱凡

Team #29831Page 2 of 29

第二届“认证杯”数学中国

数学建模国际赛 编 号 专 用 页

参赛队伍的参赛队号：（请各个参赛队提前填写好）：1273

竞赛统一编号（由竞赛组委会送至评委团前编号）：

竞赛评阅编号（由竞赛评委团评阅前进行编号）：

Team #29831Page 3 of 29

Risk Management Based on Modeling Risk

in Economic System

Abstract：

According to the target of economic system, the process of value realization for the economic activities in an economic system was analyzed.

Then the definition of ―risk‖ in such economic systems was provided , and the impact of its components:―risk events‖,―risk factors‖ , and ―risk genes‖ was analyzed. Based on the relations of risk potential and risk events, the high-level risk model was developed.

Furthermore, based on the relation of ―risk factors‖ and ―risk genes‖, the model for analyzing low-level risk in an economic system was developed. The high-level and low-level models are coordinated by the interaction of risk events and risk factors. There are many factors cause a risk, pass to be engaged in to actually set out, after surveyed the basic market data and used the high-level and the low level as a framework. Also one of the most important aspect of risk --market risk are discussed .In the discussion of market risk management, volatility modeling and vortfolio risk are referred.

Furthermore, operational risk and regulation management is analyzed, using hedge funds as an example. In the discussion of this part,Basel II framework andhedge fund due diligence are referred.

With this model, the risk analysis can deal with large amount of information of different levels, while focusing on particular concerns at the same time.

Key words:

Market data, risk events, risk factors, risk genes, market risk, volatilitymodeling, portfolio risk, covariance estimation

Team #29831Page 4 of 29

Contents

1.Introduction„„„„„„„„„„„„„„„„„„„„„„„„„„3 1.1economiccalamity„„„„„„„„„„„„„„„„„„„„„3

1.2The mathematical model in the risk„„„„„„„„„„„„„„„„3 1.3 The problems of mathematical model„„„„„„„„„„„„„„„3

2. The Description of Problem„„„„„„„„„„„„„„„„„„„„4 2.1 How do we define the risk„„„„„„„„„„„„„„„„„„„„4

2.1.1The characteristics of risk„„„„„„„„„„„„„„„„4 2.1.2 From the angle of applicable object„„„„„„„„„„„„4 2.2 Originals of the financial/economic crises„„„„„„„„„„„„„5 2.3 The classification of financial risk„„„„„„„„„„„„„„„„5 2.4 Other human factorson risk„„„„„„„„„„„„„„„„„„„5 3. Models„„„„„„„„„„„„„„„„„„„„„„„„„„„„6 3.1 Basic Model. „„„„„„„„„„„„„„„„„„„„„„„„6 3.1.1 Symbols and Definitions„„„„„„„„„„„„„„„„„„6 3.1.2 Assumptions „„„„„„„„„„„„„„„„„„„„„„7 3.1.3 The Foundation of Model„„„„„„„„„„„„„„„„„„7 3.1.4 Solution and Result „„„„„„„„„„„„„„„„„„„19 3.1.5 Analysis of the Result„„„„„„„„„„„„„„„„„„„19 3.1.6 Strength and Weakness„„„„„„„„„„„„„„„„„„„20

4.Conclusions„„„„„„„„„„„„„„„„„„„„„„„„„„„20 5.References„„„„„„„„„„„„„„„„„„„„„„„„„„„ 21 6.Appendix„„„„„„„„„„„„„„„„„„„„„„„„„„„22 6.1 Statistics„„„„„„„„„„„„„„„„„„„„„„„„„„„22 6.2 letter to the financial investment firms„„„„„„„„„„„„„„„25

Team #29831Page 5 of 29

I. Introduction

In order to indicate the origin of the economic calamity , the following background is worth mentioning.

1.1 economic calamity

The global financial crisis that began in summer 2007 has continued to wreak devastating loss and damage across all markets, all economies, and all countries. An initial liquidity contraction in the financial markets was transformed into a full solvency crisis following the collapse of Lehman Brothers on September IS, 2008. This crisis was almost wholly unpredicted and led to a massive collapse in growth and investment across the world. Leading emerging markets were not immune as any possible decoupling or separation evaporated. [1Financial Crisis-U.K. Policy and Regulatory ResponseWalker, George A. The International Lawyer44.2]Ineffective supervision, questionable investments and poor operations of financial institutions, and in so doing highlight the need for a new financial regulatory framework that willaddress these problems and rebuild thepublic’s confidence in the financial industry asa whole.

1.2The mathematical model in the risk

The 2007 credit crisis was a wake-up call with respect to model usage. It has been alleged that the misuse of risk management models helped to generate the crisis.

The models were supposed to simulate the complex interactions of many market forces on one another, including fluctuations in markets, changing interest rates, prices of various stocks, bonds, options and other financial instruments. Even if they did that--that's arguable--they failed to account for one important scenario: What happens when everybody wants to sell all their holdings at the same time? This is precisely what happened in those dark days of September 2008, when the U.S. government decided not to bail out Lehman Brothers, and the venerable institution defaulted on its creditors. The domino effect of collapse was averted only by massive infusions of money from the federal government.

1.3The problems of mathematical model

Through the risk models indicated that the chance of any major institution defaulting was minimal. A big problem was that the models omitted a major variable affecting the health of a portfolio: liquidity, or the ability of a market to match buyers and sellers. A missing key variable is a big deal--an equation that predicts an airplane flight's risk of arriving late will not be very reliable if it has no mathematical term representing weather delays.

There are multiple reasons for a state of the economy. Blaming the economic calamity on risk models would be an oversimplification. There are other human factors--political and regulator ones--certainly came into play.

Team #29831Page 6 of 29

II. The Description of the Problem

2.1How do we define the risk

The concept of risk is an outgrowth of our society's great concern about coping with the dangers of modern life. Risk refers to the possibility of an event about what we don't want to consequences.

1) The characteristics of risk

Financial markets are becoming increasingly sophisticated in pricing , isolating , repackaging,and transferring risks. Tools such as derivatives and securitization contribute to this process, but they pose their own risks.We are seeking a general definition. Any general definition must compass all of situations.The situations may appear disparate, but they share certain common elements.First,people care about the outcomes. If someone has a personal interest in what transpires,that person is exposed. Second, people don’t know what will happen.In each situation,the outcome is uncertain,It seems that risk entails two essential components and some smaller aspect :

 Objectivity  Controllability  Identifiability  Accumulation  Loss

 The likelihood of the events  the uncertainty of occurrence time  the uncertainty of the result

Based on the above, one can draw a conclusion that if the risk is uncertain, the result of a risk may lead to lost profit,.like financial risk Belong.

2)From the angle of applicable object

Risk is a condition of individuals--humans and animals--that are self-aware(3). Organizations,companies,and governments are conduits though which individuals--members, investors,employees,and such--take risk.Looking through a company to see who ultimately bears specific risks can be enlightening.For example,increasing the accountability of managers increases career risk for those managers but tends to reduce price risk for stockholders.

2.2 Originals of the financial/economic crises

There are many originals of economic calamity, economic policy mistakes,raw materials nervous,natural disasters,the consequences of globalization.The financial policy errors . From the picture 1, originals of the financial/economic crises is simple to understand.

Team #29831Page 7 of 29

Picture 1 origin of the financial/economics crises

2.3The classification of financial risk

In general, factors of financial risk can be divided into two parts:

 Objective factors. For financial institutions, this risk is objective existence and not transfer

with the willing of the people. Including country risk, interest rate risk, exchange rate risk, policy risk, liquidity risk, credit risk, investment risk, etc., most risks caused by non-artificial factors;

 Subjective factors, such as management risk, etc. The reasons are that the salesman make

some mistakes inadvertently, mismanagement, lax supervision. These problems are inevitable to some extent .

2.4 Other human factorson risk

Unfortunately, missing illiquidity risk wasn't the only major problem. Financial risk models have been designed to focus on the risk faced by an individual institution. That always seemed to make sense because institutions are concerned only with their own risk, and regulators assumed that if the risk to each individual institution is low, then the system is safe. But the assumption turned out to be poor, says Rama Cont, director of Columbia University's Center for Financial Engineering. In a system where many interdependent components each have a low risk of failure, he notes, systemic risk can still be excessive. Imagine 30 people walking side by side across a field with their arms around one another's shoulders--any one person may be unlikely to stumble, but there's

Team #29831Page 8 of 29

a decent chance someone in the group will, and that one stumbler could bring down a chunk of the line. That's the situation financial institutions are in, Cont says. \"Up through 2008, regulators weren't considering the connections between these banks in assessing risk,\" he observes. \"They should have at least noticed that were all highly invested in the subprime mortgage market.\"

III. Models

3.1 Basic Model

High-level risk model Low-level risk model Volatility model

3.2Terms, Definitions and Symbols

Symbols Meaning Risk Value Symbols Meaning Environmental Parameter Environmental Effect Function for Rv Re ERG VRGC[i] FI REC REQ Risk Events RGC[i] screening value Risk Events Character Risk Event Qualification risk factor value Rf RGM RGC RG Risk Factors Risk Gene Measure Risk Gene Character Risk Gene Self-Growth function RFV FSG t time 3.3 Assumptions

 Risk model in this paper, the economic system is composed of value at risk, risk

events, risk factors and risk factors of the risk factors of four layer structure.

 A specific risk event is caused by a variety of risk factors, and there is a dominant risk

factor in the event.

 Risk factors are made up of multiple risk factors, and there is an effect of interaction

between them.

3.4 The Foundation of Model and Solutions

3.4.1 The high-level model

The economic system in the realization of system function events occurred in the process of economic activities and the result has certain uncertainty. This kind of

Team #29831Page 9 of 29

uncertain structure into a risk of the system. Defined for the system of risk events like this! In a certain period of time, Risk events comprehensive constitute system risk .so defined Systematic risk is as follows;

RvE[Re]

In this formula, E[X] is the expectations in X.Rv represents systematic risk value,

Rerepresents research period risk events by the profit and loss value (which is a Risk

value is within the system between risk events in the system in this time of the profits and losses of the system in terms of target value).In the presence of system over a period of time A risk event .The formula (1) expressed as formula (2).

RvE[Rei]E[Re]E[Re]...E[Ren]

12The establishment of Rei linear superposition is ensured by each risk event’s independence.

In order to determine whether Reapply to Rvor determine the degree of Re role in

Rv corresponding to a particular risk event. It’s necessary to introduce the discrimination function DSF (t, d) with parameter time t and degree d as follows:

DSF(t,d)0,tT0,d,t0,T (3)

Furthermore, the formula (2)can be represented as formula (4):

RvE[Re1]DSFRe1(tRe1dRe1)E[Re2]DSFRe2(tRe2dRe2)...E[Ren]DSFRen(tRendRen) (4)

In the expression of formula 4, each risk events are independent, i.e. there is no

coupling between risk events or other correlation. while in practice, more or less，there is a link risk between events. Before applying formula (4), it’s necessary to decouple event risk analysis and processing. Risk events of decoupling, is the most important purpose of risk analysis. In the following section, the paper puts forward to analyze risk factors and risk factors. And the relationship between risk events to as decoupling analysis, the method of risk events and event risk constitute the underlying model is given

3.4.2 The underlying model constituted by risk events and risk decoupling

analysis Based on the definition of risk, risk events occur and can lead to the results of the risk of degree is decided by the nature of the risk factors and its measure interaction,that is a reflection of the necessity of risk characteristics. So from the beginning of the risk factors interaction layer, there is a risk events coupling phenomenon; Risk factors for itself, its form is a variety of factors, and there are interactions between the elements

Team #29831Page 10 of 29

of a risk factor of the phenomenon. In the introduction of the concept of risk factors, a term used to describe the elements of risk factors, risk factors and risk factors are defined as follows: the relationship between risk factors is composed of a variety of different nature of the risk factors, the nature of the risk factors by the form factor for subset in the common nature of specific time and environment, individual risk factors do not constitute a risk factor.

Define Risk Gene (referred to as RG ) includes the following content :

Risk Gene Character(referred to as RGC): in the process of genes constituting factors, the factor’s character in which the gene plays a leading role. The risk gene character can be determined according to the factors within the designated event risk scope : a gene can have multiple characters, namely factor can contain a variety of properties, the characters of one kind of genes contain common character genes of this kind all have.

Risk Gene Measure(referred to as RGM):represent the role of a property in size in the process of genes constitute factors, the measurement of different risk genes can be compared.

Definition: RG(RGC[i],RGM[i]),(i1,...,n).

The following features and relationship exist between gene and gene:

(1) Growth characteristics: the gene changes along with the change of measure value

of one or more genes. To describe the growth characteristics, introduced the

growth function FSG: RGMt1g(RGMt)FSGRGMtT （5） DecomposeFSG,For the relationship between RGM[i] of a certain risk gene character RGC[i] in the gene factor and

RGM[k],k{1,2,...,i1,i+1，i+2,...,n} , except forRGM[i]:

Team #29831Page 11 of 29

RGM[i]t1RGMtf(RGM[k]),k(1,2,...,i1,i1,...n).FSG(1,i)..FSG(i1,i)RGM[i]t1[RMG[1]t,...,RMG[i1]t,RMG[i1]t,...,RMG[n]t].FSG(i1,i)..FSG(n,i).FSG(1,i)..FSG(i1,i)RMGtTFSG[i]RMGtT1FSG(i1,i)...FSG(n,i)(6)

Conclusion:FSG[FSG[1],FSG[2],...,FSG[n]]T

(2) Environmental effect characteristic: describe the function group that describes

how much gene’s growth was affected due to the impact of external

environment .According to the gene’s dependence on the environment, there are multiple function together in a certain combination way at the same time. To describe the environment’s impact on the gene character and gene measure, introduce concepts as follows: environment parameters, environment effect function and environment function set.

Environmental impact on the economic system of this kind of open system is the important factor of the system state change, and the existence of correlation between the scope of environmental impact and environmental complexity determines that a quantitative evaluation about the impact on environment is difficult to undergo. In the research on the risk of economic system, the environment has multi-dimensional complex impact on risk factors. On the

analysis of the specific risk factors under the influence of the environment, it can be found that the main environmental impact factors are limited, and the

complexity of correlation in the theoretical research degree is acceptable. Based on the characteristics, when establish the model that reflect the impact on the risk genes, the definition of the environmental parameter and environmental effect function can be based on individual risk gene.

For RG, define the corresponding Environmental Parameter

ERG[ERG(RGC[i])]

Environmental Effect Function for RGC[i] is

VRGC[i]vRGC[i](ERG(RGC[i],RGM[i])

Team #29831Page 12 of 29

Then, RGM[i]t1VRGC[i]vRGC[i](ERG(RGC[i],RGM[i]t) With a comprehensive analysis, after risk genes (RG) were affected by environment, their impacts on RGM can be represented as follows:

RGC[1]tVVRGC[2]t...VRGC[n]tRGC[1](ERG(RGC[i],RGM[1]t)vvRGC[2](ERG(RGC[i],RGM[2]t)...,i1,...,n （7）

v(E(RGC[i],RGM[n])RGC[n]RGtRGM[i]t1Here, vRGC[i](ERG(RGC[i],RGM[i]t)constitute environmental effect function set,the collection of the function can be linear, also can be non-linear. For the environmental

effects of nonlinear function, it can be linearized processing. On

vRGC[i](ERG(RGC[i],RGM[i]t) polynomial expansion withRGM[i]

extractRGMtfrom the right of the formula (6),so the formula(6)can be represented as:

RGMt1VRGMtT (8)

Combine formula (8)with formula(5),

RGMt1VFSGRGMtT________ (9)

The formula (9), FSGrepresents the multipliable matrix transformation ofFSG

(3) Gene interaction characteristics: between the genes which constitute the factor,

there exists the interaction phenomenon and there are three important forms of interaction as follows: weaken mode, exclusive mode, strengthen mode. The weaken mode behaviors as that when Gene A and Gene B of different characters exist at the same time, also, the gene measure of A is bigger than that of B, then, the gene measure of A decreases while that of B remain the same, and vice versa. The exclusive mode behaviors as that when Gene A exists, if Gene B exists and the gene measure of A is bigger than that of B, then the B doesn’t exist, if the gene measure of A is smaller than that of B, then A remain the same and Gene B can be represented as A.The strengthen mode behaviors as that when Gene A and Gene B of different characters exist at the same time, also, the gene measure of A is bigger than that of B, then, the gene measure of A remain the same while that of B decreases, and vice versa.

Correspond with the mode of interaction, define three kinds of action function: The weaken function——W(RGA,RGB)

Team #29831Page 13 of 29

The exclusive function——O(RGA,RGB) The strengthen function——I(RGA,RGB)

The constitution of factors has screening impact on gene characters, the principle of screening reflected the chosen weight value when gene measure constitute the risk factor value(RFV).Define the risk factor value RFVV(RG,Q,FI),in the formula, RG is the value of risk gene in the risk factor, Q is the comment value which RFV gives to RG. Boolean parameter FI is the screening value which RF gives to RG. Then, RFV(Qj(RGMiFIj,i)) (10)

j1i1nmIn the formula (10), the number of risk gene that risk value RF contains is n and the number of gene character of risk gene is at most m. Risk factor value is constituted by weighed gene measure. Considering more complex risk factors,the correlation

between risk measure and risk factor value need to be represented by more complex function.

Define the expected risk events which concuding Risk Events Character(REC)、Risk Event Qualification(REC),according to the definition, it can be concluded that the most valuable characters of risk events are event time, event properties and the reference result of the degree of events’ influence. The characters are present as follows:

If REQ[i]RFV[REC[i]]0then the risk event whose risk event character is

REC[i] will happen, and the risk factor value is RFV[REC[i]], according to the RGMt0 of current risk gene, calculate ti1 when REQ[i]RFV[REC[i]]0 , ti1

represents the expected time when risk event occur.

According to REQ[i]RFV[REC[i]]0(i1,n),the value ofRFV[REC[i]] ,

t[REC[i]],introduced risk event evaluate values and the evaluation function

REEr(REQ,REV[REC],t[REC]),then Reiand tReiin the formula(4) can be

represented as follows:

ReikiREEkir(REQi,REV[RECi],t[RECi]) (11)

Team #29831Page 14 of 29

tReit[RECi] (12)

From what has been discussed above, the risk value (Rv) of the system can be completely represented by the formula (4)-(7) and (9)-(12). 3.4.3 Volatility modeling in market risk management

Volatility – especially its forward-looking estimate – is the central issue for market risk management. There are two broad approaches to extracting volatility estimates from returns data: historical and implied. Remarkably, while the level of returns is very difficult to predict from historical data due to market efficiency forces, the expected volatility is easier to forecast due to the volatility effect. As the frequency of trading has increased over time and data have become available, the estimation of historical volatility has moved to intraday time scales, where it is commonly known as realized volatility.

Implied volatility exploits the fact that the value of put and call options is positively related to volatility. Except for volatility, all of the other inputs to the basic Black-Scholes pricing model are unambiguously observable. Conditioning on these other inputs (i.e., strike price, time to expiry, risk-free interest rate, and underlying price), the formula for option value becomes an invertible function of the volatility, and one can back out the model-implied volatility estimate for a given market price for the option. Because they are based on current (i.e., forward-looking) option prices, implied volatilities have an advantage over historical volatilities in adapting to abrupt shifts between high-volatility and low-volatility regimes.

3.4.4 Portfolio risk and covariance estimation in market risk management

The volatility, measured as standard deviation of returns, for a portfolio is a simple function of the covariance matrix and the fractions of the portfolio invested in each stock. Small (or negative) correlations increase diversification benefits, making idiosyncratic stocks especially attractive for diversified portfolios.

―Market depth‖ refers to the number of potential buyers or sellers willing and able to take the other side of a trade. Market depth need not be infinite to create a liquid market, but merely adequate relative to the net order flow it must absorb. Figure YYY depicts a market with asymmetric depth on the bid and ask sides.

Team #29831Page 15 of 29

Picture 2: Market Depth in a simple limit order book

Covariance estimation also usefully illustrates many of the challenges of working with actual financial data. The covariance matrix is a theoretical ideal; in practice, real data are messy and uncooperative, and a number of complications can arise (see Alexander, 2008, ch. II.3 for an overview of the basics). To keep the problem manageable, it is common to impose some structure to reduce the number of parameters, for example assuming that all correlations come from shared exposures to a limited set of underlying factors, instead of allowing every stock to correlate separately with every other (Briner and Connor, 2008). Another problem is that many popular multivariate techniques require special restrictions to guarantee that the estimated covariance matrix is positive semidefinite.

Structural models of correlation are not merely expedients for estimation of large systems; they also typically have useful interpretations in terms of other markets and fundamentals, thus facilitating hedging strategies as well as the understanding and communication of risk exposures. The most basic and best known of these structural models posits that there is a broad-based ―market‖ factor,RM,responsible for all co-movements in returns, and that any other variability,i, (which may be significant) is firm-specific

Ri i  ßiRM i (1)

The market factor here is typically proxied by a standard market index, such as the S&P 500 (S&P, 2012b), with the parameters αand ßestimated by regression analysis.

i

i

The firm-specific error term,i,typically neither hedged nor priced because it can be

Team #29831Page 16 of 29

diversified away. The parameter ßicaptures the stock's sensitivity to the general market factor. This linear sensitivity might be hedged as appropriate, for example via an offsetting position in stock index futures. Given N stocks, notice that the estimation of a full NNcovariance matrix effectively collapses here to the estimation of the 2Nα and ß parameters (along with estimates of means and variances for RM M

+ i);

when N is large, this parsimony is enormous.

The so-called ―market model‖ in equation (1) is similar to the famous capital asset pricing model (CAPM) (Elton, Gruber, Brown and Goetzmann, 2009). Both models relate individual stock returns via a sensitivity parameter (usually labeled ―beta‖) to the return on a market portfolio or market factor. The CAPM differs in two key respects: (a) it is an equilibrium model derived from first principles as to how the market should behave if the assumptions are correct; and (b) the equilibrium result – the ―security market line‖ – is stated in expectations, and in terms of the market return premium over a risk-free interest rate, Rf ：

E(Ri)-Rf=i[E(RM)Rf] (2)

Risk managers prefer the simpler and more flexible market model (1), as risks must still be managed whether the market is in equilibrium or not. The market model is also easily extended to incorporate other factors; a more general form of (1) is:

Ri i  ßi1RM  ßi2RS jßi3jFj i(3)

where now there are many factors, including the market return, R, the overall

M

return on the relevant industry sector, R, and any other factors, F, that are deemed

S

j

appropriate, such as ―momentum‖ or firm size. Again, the sensitivities are linear, which facilitates hedging, and provides a natural economic interpretation, which facilitates risk reporting and strategy development. 3.4.5 Operational risk: the example of hedge funds

Operational risk is a wide-ranging category, encompassing hazards as diverse as fraud, software bugs, and natural disasters. Roughly speaking, there are two general approaches to operational risk measurement: one focuses on processes, while the other focuses on outcomes. These two general philosophies both have strengths and weaknesses, and they are not mutually exclusive. We offer examples of two frameworks that illustrate the two ends of this continuum, although neither is confined to one extreme. The Basel guidance, with its focus on capital requirements, tends to emphasize the realized outcomes and the estimation of a loss distribution from operational events. The framework of the Committee of Sponsoring Organizations of the Treadway Commission (COSO) emphasizes operational processes and the need for well designed controls; it is a largely non-statistical approach.

Team #29831Page 17 of 29

3.4.5.1 Basel II Framework: Operational Risk Capital Measurement

In its updated ―International Convergence of Capital Measurement and Capital Standards: A Revised Framework,‖ the Basel Committee on Banking Supervision (The Committee) defines operational risk as the risk of loss ―due to failed or inadequate processes, people, or systems, or from external events. The framework provides three methods to comply with the requirement to set aside operational risk capital: (1) the Basic Indicator Approach, (2) the Standardized Approach, or (3) the Advanced Measurement Approaches.

Banks must demonstrate they have effective risk management and controls in place, including active involvement from a board of directors and senior management, an operational risk system that is ―conceptually sound‖ and sufficient resources in each of the business lines as well as in the risk and control areas (―qualifying criteria‖). The bank must also demonstrate a risk management framework and regular audits of the operational risk management process and key inputs. Internationally active banks have

additional criteria if they follow this approach, including the need to demonstrate the ability to collect, analyze and report loss data across all business lines.The most visible operational risk events for financial services companies in current times are rogue trader events, but events with no publicity, such as system outages, also can easily affect a firm’s bottom line

Table 1 Basel Risk Categories

Basel Risk Categories Internal Fraud External Fraud Employment Practices and Workplace Safety Clients, Products and Business Practices Damage to Physical Assets Business Disruptions and Failures Execution, Delivery and Process Management Examples misappropriation of assets, tax evasion, intentional mismarking of positions theft of information, hacking damage, third-party theft and forgery discrimination,workers compensation, employee health and safety market manipulation, antitrust, improper trade, product defects, fiduciary breaches, account churning natural disasters, terrorism, vandalism utility disruptions, hardware failures software failures, data entry errors, accounting errors, failed mandatory reporting, negligent loss of client assets

Team #29831Page 18 of 29

3.4.5.2 Hedge Fund Due Diligence

The foregoing outlines two very different approaches to operational risk monitoring, reflecting two distinct traditions. The Basel process comes from the data- and probability-centric context of international bank capital regulation. It provides

detailed measurement strategies specific to finance. The COSO process comes from the controls- and accountability-centric context of internal auditing. It is much less focused on the statistical distribution of loss events, and more focused on processes and how they are managed. This distinction between operational risk viewpoints is similar to the contrast that Brown (2012) draws between ―bottom-up‖ (e.g., Basel) approaches and ―top-down‖ approaches based on evidence of legal and regulatory problems in operational due-diligence reports (see also Brown, et al., 2008). In this context, hedge funds represent an interesting accidental experiment. While neither framework is a requirement for hedge funds, they nonetheless face a very immediate operational risk challenge. As they compete for funds against each other and against well capitalized, regulated, and (relatively) transparent banks, they must regularly convince skeptical investors that their funds will be safely kept.

Still, large investors and fund-of-fund managers should and do insist on formal and due-diligence reviews of the operations of candidate hedge funds. These reviews

focus on fraud, conflicts of interest, incompetence, and the adequacy of processes and systems (Scharfman, 2008). They typically include a detailed pre-investment review, including background checks and interviews of key staff. Sharfman (2008)

recommends a scorecard approach to guarantee that all relevant dimensions of the review are covered and comparable across funds. On the other side, a scorecard

approach is also modeled by the Managed Funds Association (MFA, 2007) as part of its recommended practice for funds in their interactions with potential investors. Regular (e.g., annual), on-site, post-investment reviews confirm that no material deterioration in operational quality has occurred. Occasional on-site reviews can be supplemented with continuous monitoring of industry news and legal events such as court docket reports and new case filings. MFA (2009), in their best-practice recommendations, focus heavily on operational issues and processes:

Disclosure to investors

Valuation governance to mitigate model risk and misreporting

Risk management (market, credit, liquidity, operational, legal/compliance) Operational processes (trading and business)

Team #29831Page 19 of 29

Compliance, ethics and conflicts of interest Anti-money laundering compliance Business continuity and disaster recovery

Interestingly, both the internal and third-party evaluation and monitoring approaches typical

in the hedge-fund industry seem closer to the process-centric COSO approach than the loss-distribution approach of the Basel framework designed specifically for financial institutions.

3.4.6 Regulation Management , Reform and the balanced system

Government regulation has been defined as ―the diverse set of instruments by which governments set requirements on enterprises and citizens‖. From the public finance perspective, governments develop regulatory policies to address problems of market failure and new concerns about social values. Public administration scholars need to understand not only the causes and remedies of the financial crisis, but also the array of designs, methods, and management arrangements of the financial regulatory system (Khademian, 2009). It is important to improve regulation management by enhancing administrative capacity and our understanding of the complexities of the regulatory system.

We don't give too much power to regulators because we know two things: (1) they will abuse it, and (2) they will be late to the game, as regulators always are. For this reason we rely on a market trigger-namely, the CDS price of an institution's junior long-term debt. When that price reaches a certain threshold, regulators should intervene-make a margin call- and wind down the failed institution. We don't want to abuse the market trigger either, because we are afraid of what are called \"self-fulfilling prophecies\"-the market gets very worried, regulators intervene, penalize the debt, and this fulfills the prophecy. For this reason we have introduced the stress test as a circuit breaker to prevent this escalation. This circuit breaker, however, can create perverse incentives for the regulators. So we need a system to penalize them if they make mistakes-namely, the loss of their investment in junior long-term debt.

This system should apply only to very large financial institutions, because other institutions can and do fail, subjecting them to the normal market discipline. Will it be costly? Yes, it will, and it should be costly. It should be costly to undo a major distortion that now exists, a distortion that favors large institutions at the cost of small institutions. Today the implicit too-big-to-fail doctrine is a subsidy to large financial institutions, with a lot of negative effects.

In particular, there is more concentration in the financial sector, which is bad for

Team #29831Page 20 of 29

consumers and taxpayers. More institutions will have to be bailed out in the future. Moving to a market-based regulatory regime would remove the too-big-to-fail bias and reintroduce a fair marketplace

3.5 Analysis of the Result

The high-level and low-level models are coordinated by the interaction of risk events and risk factors. There are many factors cause a risk, pass to be engaged in to actually set out, after surveyed the basic market data and used the high-level and the low level as a framework, most important aspects of market risk -- Volatility modeling and Portfolio riskare discussed. With these models, the risk analysis can deal with large amount of information of different levels, while focusing on particular concerns at the same time.

Financial risk management is the industry's analytical response to the challenges presented by an imperfect world. The analysis requires information as its key input, which in turn requires enormous amounts of data. The output is an array of concrete decisions, such as whether to make or deny a loan, whether to add capital to the firm, or how to hedge a position.

These statistical models of financial risk are valuable and appropriate, but they are not all-encompassing. Risk managers should also give due weight to uncertainties and the tools that can address them.

Similarly, subtle organizational hazards can be often be tracked in data on operational risk events, but may often be better approached with ―softer‖ tools such as formal processes, risk controls and governance, internal and external transparency, and the alignment of risk-taking incentives. Risk managers should never forget that tools are only as good as the people that use them, and that confining oneself to the output of a model when a open-eyed view of the market would direct a different outcome is not effective risk management. Good risk management requires toolkits to handle both Knightian risk and uncertainty.

3.6 Strength and Weakness

Strength: A clear emphasis on statistical tools emerges in our overview of risk management

models and practices. This is largely an artifact of a modern financial culture, which is awash in data and heavily informed by a deeply rooted mindset of empirical and statistical models of markets.

Weakness: Although these models can capture inter-temporal clustering, they are challenged

by abrupt changes in the statistical regime.

IV. Conclusions of the problem

Financial risk management is the industry's analytical response to the challenges

Team #29831Page 21 of 29

presented by an imperfect world. The analysis requires information as its key input, which in turn requires enormous amounts of data. The output is an array of concrete decisions, such as whether to make or deny a loan, whether to add capital to the firm, or how to hedge a position. Knight (1921) introduced a simple but fundamental classification of the information challenges we face: between Knightian risks, which can be successfully addressed with statistical tools; and Knightian uncertainties, which cannot. A clear emphasis on statistical tools emerges in our overview of riskmanagement models and practices. This is largely an artifact of a modern financial culture, which is awash in data and heavily informed by a deeply rooted mindset of empirical and statistical models of markets (Mackenzie, 2008).

These statistical models of financial risk are valuable and appropriate, but they are not all-encompassing. Risk managers should also give due weight to uncertainties and the tools that can address them. For example, there are statistical tools to apply extreme value theory to rare events, but these may often be better approached using scenario analysis and stress testing. Similarly, subtle organizational hazards can be often be tracked in data on operational risk events, but may often be better approached with ―softer‖ tools such as formal processes, risk controls and governance, internal and external transparency, and the alignment of risk-taking incentives. Risk managers should never forget that tools are only as good as the people that use them, and that confining oneself to the output of a model when a open-eyed view of the market would direct a different outcome is not effective risk management. Good risk management requires toolkits to handle both Knightian risk and uncertainty.

In summary, this whirlwind tour of risk management information generates two broad lessons. The first is simply that information matters. While there will always be a large component of randomness in the world – and especially in financial markets – there are also areas of ignorance that can be illuminated simply by applying oneself thoughtfully and diligently to better understand the data. To paraphrase Arnold Palmer, the more you inform yourself, the luckier you will get. Where data are available, statistical tools will likely help, but the absence of high-quality numeric data does not imply the absence of potential hazards. Second, risk management is an enormous and enormously intricate topic. The long list of topics mentioned here is merely the tip of an iceberg. We hope the references provided will serve as useful pointers to more detailed investigations.

Team #29831Page 22 of 29

V. References

【1】The Science of Bubbles and Busts. Gary Stix in Scientific American, Vol.

301, No. 1, pages 78-85; July 2009.

【2】The Failure of Risk Management: Why It's Broken and How to Fix It. Douglas

W. Hubbard. Wiley, 2009.

【3】Too Big to Fail: The Inside Story of How Wall Street and Washington 【4】Fought to Save the Financial System--and Themselves. Andrew Ross Sorkin. Penguin, 2010.

【5】Risk Management Models: Construction, Testing, Usage. Robert A. Jarrow in Journal of Derivatives. Summer 2011, Vol. 18, No. 4, pages -98; Summer 2011. 【6】Improving Financial and Regulatory Management.Liou, Kuotsai Tom1

Dec2013, Vol. 37 Issue 2, p199-207. 9p.

【7】Financial Crisis-U.K. Policy and Regulatory Response

Walker, George AThe International Lawyer44.2 (Summer 2010):

751-7.

【8】A Multiperiod Bank Run Model for Liquidity Risk* Gechun Liang1, Eva Lütkebohmert2 and Yajun Xiao2 【9】Integrating multi-market risk models

Shepard, Peter GThe Journal of Risk10.2 (Winter 2007/2008): 25-45. 【10】Quantitative risk analysis offshore—Human and organizational factors

Jhon Espen SkogdalenJan Erik Vinnem

【11】A FORMULA FOR ECONOMIC CALAMITY Freedman, David H.,

Scientific American, 00368733, Nov2011, 卷 305, 发行 5

【12】Addy, Christopher (2008), ―Operational Due Diligence for Hedge Funds,‖ presented at the Public Pension Financial Forum.

【13】Alexander, Carol, 2008, Market Risk Analysis, Volume II: Practical Financial

Econometrics, Wiley Finance________, 2009, Market Risk Analysis, Volume IV:

Team #29831Page 23 of 29

Value At Risk Models, Wiley Finance.

【14】Alexander, Carol, and Elizabeth Sheedy, 2008, ―Developing a stress testing f ramework based on market risk models,‖ Journal of Banking and Finance, 32, 2220–2236.

【15】Allen, Linda, Gayle DeLong, and Anthony Saunders, 2004, ―Issues in the credit risk modeling of retail markets,‖ Journal of Banking and Finance, 28(4), April, 727-752.

【16】Altman, Edward, 1968, Financial Ratios, Discriminant Analysis and the

Prediction of Corporate Bankruptcy, Journal of Finance,23(4),September, 5-609.

VI. Appendix

6.1 Statistics

Team #29831Page 24 of 29

Table 2: Standardized Approach - Business Line Mapping Beta Level 2 Activity Groups Level 1 Facto rs Corporate 18% Corporate finance Mergers and acquisitions, finance underwriting, privatisations, Municipal/government securitisation, research, debt finance (government, high yield), equity, Merchant banking syndications, IPO, secondary private Advisory services placements Trading and 18% Sales Fixed income, equity, foreign sales exchanges, commodities, credit, Market making funding, own position securities, Proprietary positions lending and repos, brokerage, debt, Treasury prime brokerage Retail banking 12% Retail banking Retail lending and deposits, banking services, trust and estates Private banking Private lending and deposits, banking services, trust and estates, investment advice Card services Merchant/commercial/corporate cards, private labels and retail Commercial 15% Commercial banking Project finance, real estate, export banking finance, trade finance, factoring, leasing, lending, guarantees, bills of exchange Payment and 18% External clients Payments and collections, funds settlement transfer, clearing and settlement Agency 15% Custody Escrow, depository receipts, services securities lending (customers) corporate actions Corporate agency Issuer and paying agents Corporate trust Asset 12% Discretionary fund Pooled, segregated, retail, management management institutional, closed, open, private equity Non-discretionary fund Pooled, segregated, retail, management institutional, closed, open Retail 12% Retail brokerage Execution and full service brokerage Picture 3: Portfolio Loss Distributions

Team #29831Page 25 of 29

Table 3: Example of Internal Ratings Buckets Internal category PD (%) LGD (%) 1 Extremely low risk 2 Low risk 3 Moderate risk 4 Average risk 5 Acceptable risk 6 Borderline risk 7 Special cases 8 Substandard 9 Extremely high risk

EL (%) 0.000 0.025 0.075 0.250 0.750 1.500 5.000 15.000 25.000 0.00 0.10 0.30 1.00 3.00 6.00 20.00 60.00 100.00 25 25 25 25 25 25 25 25 25 Team #29831Page 26 of 29

Table4 TOTAL FINRA MEMBER FIRMS

INCOME STATEMENT & SELECTED ITEMS

Table5 Outstanding Money Market Instrument

Team #29831Page 27 of 29

Picture 4 Model risk management framework

6.2 A letter to the financial investment firms

The best way to understand models and their use is to consider an analogy.

Models are analogous to medical prescription drugs. Prescription drugs havegreat medical bene.ts if used properly, with educated use. If used wrongly, how-ever, prescription drugs can have negative consequences, even death. Becauseprescription drugs can cause death if used wrong, this does not mean that weshould stop using them. It does mean, however that we need educated use. Infact, in reality, prescription drugs should probably be used more because theysave and prolong lives. The same is true of models.

Financial markets have become too complex to navigate without risk management models. Determining a price - fair value - is not an issue because inmany cases expert judgment can provide reasonable estimates. But,

1、there is no way to hedge a portfolio, i.e. determine hedge ratios, withouta model,

2、there is no way to determine the probability of a loss (risk measures) with-out a model, and

3、there is no way to price a derivative in a illiquid market without a model. These issues are at the heart of risk management. Hence, .nancial risk

Team #29831Page 28 of 29

management models are here to stay.

Models if used properly, help decision making. Models if used improperly,generate bad decisions and lead to losses. This doesn.t mean that we shouldnot use models. Quite the contrary. It only means that we need more educateduse of models. Had models been properly used before the crisis, the 2007 creditcrisis, perhaps the crisis would have not occurred.

A new way predict bubble mamies：

1、AN EVOLUTIONARY PERSPECTIVE

Researchers at the Massachusetts Institute of Technology have combined several descriptions of how markets work--and borrowed from evolutionary theory--in an attempt to make better predictions about when buying and selling activity will become volatile and which investors will survive the turmoil. Their conception is called the adaptive-market hypothesis.

2、CORRELATION: ONE PRICE LEADS TO ANOTHER

A computational analysis based on the adaptive-market hypothesis tracks the degree to which price changes that occur on one day influence how much prices are altered on the next--in essence, how closely price changes are correlated. 3、IRRATIONAL MARKET

Upward price movements with a high degree of correlation imply that many investors are herding into the market and that a bubble may be forming, a trend that may be driven by an irrational belief that prices will go up indefinitely. 4、RATIONAL MARKET

After a bubble bursts and herding activity subsides, the market returns to the more \"efficient\" state modeled by classical economists; in an efficient market, investors hold independent beliefs about the direction of the market. 5、SURVIVAL OF THE FINANCIALLY FITTEST

The adaptive-market hypothesis also combines evolutionary theory with information about correlations and data related to the financial health of individual and institutional investors. This synthesis can predict who may adapt as market conditions change and who may fall by the wayside.

MFA (2009), in their best-practice recommendations, focus heavily on operational issues and processes: Disclosure to investors、Valuation governance to mitigate model risk and misreporting、Risk management (market, credit, liquidity, operational, legal/compliance)、Operational processes (trading and business)、Compliance, ethics and conflicts of interest、Anti-money laundering compliance、Business continuity and disaster recovery

Interestingly, both the internal and third-party evaluation and monitoring approaches typical in the hedge-fund industry seem closer to the process-centric

COSO approach than the loss-distribution approach of the Basel framework designed specifically for financial institutions.

Team #29831Page 29 of 29

Financial risk management is the industry's analytical response to the challenges presented by an imperfect world. The analysis requires information as its key input, which in turn requires enormous amounts of data. A clear emphasis on statistical tools emerges in our overview of risk management models and practices. These statistical models of financial risk are valuable and appropriate, but they are not

all-encompassing. Risk managers should also give due weight to uncertainties and the tools that can address them. For example, there are statistical tools to apply extreme value theory to rare events, but these may often be better approached using scenario analysis and stress testing.

In summary, this whirlwind tour of risk management information generates two broad lessons. The first is simply that information matters. While there will always be a large component of randomness in the world – and especially in financial markets – there are also areas of ignorance that can be illuminated simply by applying oneself thoughtfully and diligently to better understand the data. To paraphrase Arnold Palmer, the more you inform yourself, the luckier you will get. Where data are available,

statistical tools will likely help, but the absence of high-quality numeric data does not imply the absence of potential hazards. Second, risk management is an enormous and enormously intricate topic. The long list of topics mentioned here is merely the tip of an iceberg. We hope the references provided will serve as useful pointers to more detailed investigations.