1. bookAHEAD OF PRINT
Dettagli della rivista
License
Formato
Rivista
eISSN
2444-8656
Prima pubblicazione
01 Jan 2016
Frequenza di pubblicazione
2 volte all'anno
Lingue
Inglese
Accesso libero

Nonlinear Differential Equations in Preventing Financial Risks

Pubblicato online: 15 Jul 2022
Volume & Edizione: AHEAD OF PRINT
Pagine: -
Ricevuto: 16 Feb 2022
Accettato: 18 Apr 2022
Dettagli della rivista
License
Formato
Rivista
eISSN
2444-8656
Prima pubblicazione
01 Jan 2016
Frequenza di pubblicazione
2 volte all'anno
Lingue
Inglese
Introduction

In recent years, the default risks faced by financial institutions have been increasing, and the corresponding capital requirements have also been increasing. The financial market is full of risks. There are numerous examples of losses or company failures caused by poor risk management in the market. The Association of Interbank Money Market Dealers issued guidelines on credit risk mitigation tools in October 2010. This marks the official launch of the Chinese version of the CDS credit derivatives market. This favorably promotes the release of credit risk. The pricing of credit default swaps is a kind of financial credit derivatives. This allows the credit risk of risky assets to be separated from asset ownership. The market-oriented arrangement for transferring credit risk gives risky assets liquidity characteristics [1]. It has played an important role in bank credit portfolio management and the financial risk control of listed companies. Some scholars have studied the pricing of corporate bonds under the reduced model where the interest rate obeys the Vasicek model, and the default intensity is a linear function of the interest rate. Some scholars focus on the random recovery rate model. They treat the recovery rate as an exponential function of default strength. We established a pricing model and used the technique of solving partial differential equations to give an analytical solution to corporate bonds. This way, the impact of recovery parameters and default strength parameters on credit spreads is analyzed [2]. The above documents assume that the company's asset value Vt meets the standard Brownian motion-driven stochastic differential equation in a given probability space. However, the standard Brownian motion cannot describe the fractal characteristics of financial markets, such as self-similarity and long memory. Fractional Brownian motion has become the easiest way to compensate for the shortcomings of the Brownian motion model because fractional Brownian motion is a Gaussian process.

Its properties mainly include additive invariance, self-similarity, thick tail, discontinuity, and long-term correlation. This makes the fractional Brownian motion neither a Markov process nor a semimartingale. This has brought huge difficulties to avoid arbitrage opportunities in the financial market under fractional Brownian motion [3]. Some scholars have proposed using the fractional Brownian jump-diffusion model to avoid the existence of arbitrage, and the results are still not ideal.

Therefore, this article attempts to study the capital valuation adjustment of the value of the derivatives transaction contract. We use the semi-copy method to design a reasonable hedging portfolio strategy [4]. The partial differential equation of valuation adjustment is derived based on the partial differential equation of the value of the derivative transaction contract and the risk-free value of the contract [5]. At the same time, we apply the Feynman-Kac formula to derive a valuation adjustment expression closer to the market from separating the capital valuation adjustment. As mentioned earlier, the calculation of the capital valuation adjustment further revised the price of the derivative contract value.

Capital hedging portfolio strategy

We use the replication method to establish a reasonable hedging portfolio strategy. This enables hedging of derivatives [6]. Using this method to hedge against the counterparty's default risk has become an important means of bank risk management. At the same time, banks are also using scientific and reasonable calculation methods and appropriate models to measure the default risk of trading contracts. This will strengthen the management of capital valuation adjustments and reduce the cost of capital. We set the price of the underlying asset stock as a Brownian motion dS = μSdt + σSdW. In this way, a hedging portfolio strategy is established. Π=δS+a1P1+a2P2+aCPC+βS+βCX1 \Pi = \delta S + {a_1}{P_1} + {a_2}{P_2} + {a_C}{P_C} + {\beta _S} + {\beta _C} - {X^1}

This model assumes that the random process followed by the price of the underlying asset stock is geometric Brownian motion. The model does not consider the jump behavior of stock prices [7]. This article sets the price process of the underlying asset as a jump-diffusion process in the subsequent model asset assumptions. The purpose is to enable the model to cover this risk factor. At the same time, we have added stock options to the hedging portfolio. The purpose is to hedge against the volatility risk of stock prices.

Assume that the value of the transaction contract between the bank and its counterparty is V^=V^(t,S,JB,JC) \hat V = \hat V\left({t,S,{J_B},{J_C}} \right) . The value of the corresponding transaction contract without default risk is V. Where S represents the stock price that is not affected by the bank or counterparty default. JB and JC respectively represent the indicative function of the bank and counterparty default. JB and JC are independent of each other [8]. There is a deviation between the price of any derivative transaction contract and its true value. We set it to U, then we have V^=V+U \hat V = V + U

The deviation U here means total valuation adjustment, including credit valuation adjustment (CVA), financing valuation adjustment (FVA), margin valuation adjustment (MVA), collateral (COLVA), and capital valuation adjustment (KVA). U = CVA + FVA + MVA + COLVA + KVA. The valuation adjustment is the difference between the value of the derivative contract containing credit risk, financing cost, margin, collateral and capital cost, and the value of the contract without default risk.

When the bank or counterparty defaults, the liquidation or claim of the trading contract position is usually determined by the Mark-to-Market M (t, S) of derivatives [9]. Let M (t, S) = V (t, S) determine the value at market value. When the bank defaults, there are: V^(t,S,1,0)=gB(V,X) \hat V(t,S,1,0) = {g_B}(V,X)

When the counterparty defaults, there are: V^(t,S,0,1)=gC(V,X) \hat V(t,S,0,1) = {g_C}(V,X)

gB (V, X) = [V (S, t) − X (t)]+ + RB [V (S, t) − X (t)] + X (t) in formula (3).

gC (V, X) = RC[V (S, t) − X (t)]+ +[V (S, t) − X (t)] + X (t) in formula (4).

Where RB, RC is the recovery rate of the contract value calculated by market value and the difference between the collateral.

When the value of the transaction contract between the bank and its counterparty is V^=V^(t,S,JB,JC) \hat V = \hat V\left({t,S,{J_B},{J_C}} \right) , it needs to hedge the default risk of the transaction contract [10]. According to the semi-copy method, the specific form of a hedging portfolio strategy Π that satisfies V^+Π=0 \hat V + \Pi = 0 is as follows: Π=δSS+aC(S,t)+a1P1+a2P2+aCPC+ϕK(t)+X(t)+βc+βs+βx+βk \Pi = {\delta _S}S + aC\left({S,t} \right) + {a_1}{P_1} + {a_2}{P_2} + {a_C}{P_C} + \phi K(t) + X(t) + {\beta _c} + {\beta _s} + {\beta _x} + {\beta _k}

Any hedging portfolio strategy should include at least different assets such as the subject matter, bonds issued by both parties to the transaction, collateral, capital, and cash accounts. Therefore, the risk should include the stock of δS unit price of S as the basic underlying asset. At the same time, increase the stock option item C(S, t) of the short position, namely a < 0. The purpose is to reduce the risk of volatility in stock prices [11]. The combination strategy also includes bonds issued by the bank with a1 unit price of P1 and a2 unit price of P2. At the same time, it also covers the aC unit price PC bonds issued by the counterparty. When the counterparty defaults, it will cause bank losses. The bank can “borrow” bond PC to hedge against losses caused by the counterparty's default. So take a short bond position PC . aC < 0. In addition, the strategy also includes the capital cost K (t) of ϕ units. One unit of collateral X (t) and repurchase cash accounts with different assets as collateral are included in the cost of different assets. βc, βS, βX and βK are the cash accounts that supply bonds PC, stocks S and stock options C(S, t), collateral, and capital costs, respectively. We explain the dynamic equations of different assets in the strategy. The stock price obeys the jump-diffusion process. This can reflect the actual market situation: dS=(μλv)Sdt+σSdW+(Y1)SdJ dS = \left({\mu - \lambda v} \right)Sdt + \sigma SdW + (Y - 1)SdJ

μ represents the expected return rate of the stock. W represents geometric Brownian motion. σ represents the volatility of the stock. The variable J represents the number of jumps of the underlying stock in [0, T] and it obeys the Poisson jump process with the parameter λ. Y represents the rate of change in the price of the underlying stock. υ = E(Y −1) and Y and dJ are independent of each other. When the stock price obeys the jump-diffusion process, the differential form of the option value C = C(S, t) with the stock as the subject is as follows: dC(S,t)=(Ct+12σ2S22CS2)dt+CSdS=(Ct+(μλv)SCS+12σ2S22CS2)dt+σSCSdW+[C(YS,t)C(S,t)]dJ \eqalign{& dC(S,t) = \left({{{\partial C} \over {\partial t}} + {1 \over 2}{\sigma ^2}{S^2}{{{\partial ^2}C} \over {\partial {S^2}}}} \right)dt + {{\partial C} \over {\partial S}}dS \cr & = \left({{{\partial C} \over {\partial t}} + \left({\mu - \lambda v} \right)S{{\partial C} \over {\partial S}} + {1 \over 2}{\sigma ^2}{S^2}{{{\partial ^2}C} \over {\partial {S^2}}}} \right)dt + \sigma S{{\partial C} \over {\partial S}}dW + \left[{C\left({YS,t} \right) - C(S,t)} \right]dJ \cr}

The prices issued by the bank are P1 and P2 respectively. The dynamic equations satisfied by bonds with recovery rates of P1 and P2 (P1P2) : dP1=r1P1dt(1R1)P1dJB d{P_1} = {r_1}{P_1}dt - (1 - {R_1}){P_1}d{J_B} dP2=r2P2dt(1R2)P2dJB d{P_2} = {r_2}{P_2}dt - (1 - {R_2}){P_2}d{J_B}

The basis difference between bonds of different grades is not considered for the time being. We use bonds with different recovery rates to offset the balance sheet impact of derivatives. r1=r(1R1)λB {r_1} = r - (1 - {R_1}){\lambda _B} r2=r(1R2)λB {r_2} = r - (1 - {R_2}){\lambda _B}

λB is the bank's default rate. r is the risk-free interest rate. Assume that the bond recovery rate of the counterparty is zero and the following dynamic equation is satisfied: dPc=rcPcdtPcdJc d{P_c} = {r_c}{P_c}dt - {P_c}d{J_c}

The repurchase account βS with stocks as collateral satisfies: dβS=δ(γSqS)Sdt d{\beta _S} = \delta \left({{\gamma _S} - {q_S}} \right)Sdt

qS represents the determined dividend yield of stock assets. γS = r + Sβ,S, Sβ,S represents the spread that exceeds the risk-free rate of return r. The repurchase account βC with bond Pc as collateral satisfies: dβC=aCqCPCdt d{\beta _C} = - {a_C}{q_C}{P_C}dt

qc represents the yield of bond Pc. When the borrowing rate of bond Pc is rc, the counterparty's default rate λc is: λC=rCqC {\lambda _C} = {r_C} - {q_C}

Collateral X (t) and its repurchase account βX satisfy: dX(t)=rX(t)dt dX(t) = - rX(t)dt dβX=rXXdt d{\beta _X} = {r_X}Xdt γX=rXr {\gamma _X} = {r_X} - r

Capital K (t) and its repurchase account βK satisfy: dK(t)=rKdt dK(t) = rKdt dβK=rKK(t)dt d{\beta _K} = - {r_K}K(t)dt dβK=rKK(t)dt d{\beta _K} = - {r_K}K(t)dt

In addition, the asset constraints of the hedging portfolio strategy are set to: (δ1+δ2)S+aC(S,t)+βS=0 \left({{\delta _1} + {\delta _2}} \right)S + aC\left({S,t} \right) + {\beta _S} = 0 αCPC+βC=0 {\alpha _C}{P_C} + {\beta _C} = 0

Since the hedging investment strategy Π satisfies V^+Π=0 \hat V + \Pi = 0 , the capital constraint of the contract value V^ \hat V at this time is: V^=a1P1+a2P2+X(t)+ϕK(t)+βX+βK=0 \hat V = {a_1}{P_1} + {a_2}{P_2} + X(t) + \phi K(t) + {\beta _X} + {\beta _K} = 0

Solving Partial Differential Equations

We derive the partial differential equation of the total valuation adjustment from the differential form of the hedging portfolio strategy containing stock options and the cost of capital. Then the Feynman-Kac formula derives the expression of the total valuation adjustment [12]. At the same time, we separate the capital valuation adjustment calculation expression. According to the Ito lemma of the jump-diffusion process, we find that the differential form of the contract value V^=V^(t,S,JB,JC) \hat V = \hat V\left({t,S,{J_B},{J_C}} \right) is: dV^=V^tdt+V^SdS+12σ2S22V^S2dt+ΔVBdJB+ΔVCdJC d\hat V = {{\partial \hat V} \over {\partial t}}dt + {{\partial \hat V} \over {\partial S}}dS + {1 \over 2}{\sigma ^2}{S^2}{{{\partial ^2}\hat V} \over {\partial {S^2}}}dt + \Delta {V_B}d{J_B} + \Delta {V_C}d{J_C}

ΔVB=gBV^ \Delta {V_B} = {g_B} - \hat V and ΔVC=gCV^ \Delta {V_C} = {g_C} - \hat V are the losses or gains when the bank and the counterparty each default. When the hedging portfolio strategy adopted by the bank is in the following form: Π=δSS+a1P1+a2P2+acPc+βC+βS+βX+βK+αC(S,t)+X(t)+ϕK(t) \Pi = {\delta _S}S + {a_1}{P_1} + {a_2}{P_2} + {a_c}{P_c} + {\beta _C} + {\beta _S} + {\beta _X} + {\beta _K} + \alpha C\left({S,t} \right) + X(t) + \phi K(t)

Over time, the differential form of hedging investment strategy Π is: dΠ=δSdS+a1dP1+a2dP2+acdPc+dβC+dβS+dβX+dβK+adC(S,t)+dX(t)+ϕdK(t) \eqalign{& d\Pi = {\delta _S}dS + {a_1}d{P_1} + {a_2}d{P_2} + {a_c}d{P_c} + d{\beta _C} + \cr & d{\beta _S} + d{\beta _X} + d{\beta _K} + adC(S,t) + dX(t) + \phi dK(t) \cr}

According to the assumptions, we assume δS = δ1 + δ2 and substitute δ1=V^Sδ2=aCS {\delta _1} = - {{\partial \hat V} \over {\partial S}}{\delta _2} = - a{{\partial C} \over {\partial S}} into the above formula. At the same time, we substitute formula (7) and formula (16), (17), (18), (19), (20), and (21) into formula (27) to sort out. Where P = a1P1 + a2 P2, PD = a1R1P1 + a2 R2 P2 and λC = rCqC. Then there are: d(V^+Π)=dV^+dΠ d\left({\hat V + \Pi} \right) = d\hat V + d\Pi

Reorganize the above formula 1 to get: dΠ+dV^=εhdJB+[tV^+AtV^(r+λB+λC)V^+λBgB+λCgCγXXγKK+r(X+ϕK+βX+βK)εhλB]dt d\Pi + d\hat V = {\varepsilon _h}d{J_B} + \left[\matrix{{\partial _t}\hat V + {A_t}\hat V - \left({r + {\lambda _B} + {\lambda _C}} \right)\hat V + {\lambda _B}{g_B} + {\lambda _C}{g_C} \hfill \cr - {\gamma _X}X - {\gamma _K}K + r\left({X + \phi K + {\beta _X} + {\beta _K}} \right) - {\varepsilon _h}{\lambda _B} \hfill \cr} \right]dt

In the formula At=12σ2S22S2+CtCS+12σ2S22CS2S+(γSqS)SS {A_t} = {1 \over 2}{\sigma ^2}{S^2}{{{\partial ^2}} \over {\partial {S^2}}} + {{\partial C} \over {\partial t}}{{\partial C} \over {\partial S}} + {1 \over 2}{\sigma ^2}{S^2}{{{\partial ^2}C} \over {\partial {S^2}}}{\partial \over {\partial S}} + ({\gamma _S} - {q_S})S{\partial \over {\partial S}} , γXrXr. It can be seen from the above formula that as long as the bank has not defaulted, the risk-free state of the transaction contract can be achieved by hedging the risk through the investment portfolio strategy. When the bank defaults, εh is the bondholder's profit and loss. Assume that the final return of the investment portfolio is H (S) when the transaction is terminated. At the same time, from the formula (28), it can be known that the portfolio strategy Π and the contract value V^ \hat V satisfy: dΠ+dV^=εhdJB+[tV^+AtV^(r+λB+λC)V^+λBgB+λCgCγXXγKK+r(X+ϕK+βX+βK)εhλB]dt d\Pi + d\hat V = {\varepsilon _h}d{J_B} + \left[\matrix{{\partial _t}\hat V + {A_t}\hat V - \left({r + {\lambda _B} + {\lambda _C}} \right)\hat V + {\lambda _B}{g_B} + {\lambda _C}{g_C} \hfill \cr - {\gamma _X}X - {\gamma _K}K + r\left({X + \phi K + {\beta _X} + {\beta _K}} \right) - {\varepsilon _h}{\lambda _B} \hfill \cr} \right]dt

Banks take into account the interests of shareholders, so they hope that the hedging strategy in the above formula can be transformed into a form of self-financing. The drift term in the above formula is required to be zero, then the partial differential equation about the transaction contract value V^ \hat V is: {tV^+AtV^(r+λB+λC)V^+λBgB+λCgC+γXXγKK+r(X+ϕK+βX+βK)εhλB=0V^(T,S)=H(S) \left\{\matrix{{\partial _t}\hat V + {A_t}\hat V - \left({r + {\lambda _B} + {\lambda _C}} \right)\hat V + {\lambda _B}{g_B} + {\lambda _C}{g_C} + {\gamma _X}X \hfill \cr - {\gamma _K}K + r\left({X + \phi K + {\beta _X} + {\beta _K}} \right) - {\varepsilon _h}{\lambda _B} = 0 \hfill \cr \hat V\left({T,S} \right) = H(S) \hfill \cr} \right.

After finishing: {tV^+AtV^(r+λB+λC)V^=γXX+γKKλCgCγBgBr(X+ϕK+βX+βK)+λBεhV^(T,S)=H(S) \left\{\matrix{{\partial _t}\hat V + {A_t}\hat V - \left({r + {\lambda _B} + {\lambda _C}} \right)\hat V = - {\gamma _X}X + {\gamma _K}K - {\lambda _C}{g_C} - \hfill \cr {\gamma _B}{g_B} - r\left({X + \phi K + {\beta _X} + {\beta _K}} \right) + {\lambda _B}{\varepsilon _h} \hfill \cr \hat V\left({T,S} \right) = H(S) \hfill \cr} \right.

The risk-free value V of the transaction contract calculated by the classic Black-Scholes-Merton model satisfies the standard BlackScholes partial differential equation: {tV+AtVrV=0V(T,S)=H(S) \left\{\matrix{{\partial _t}V + {A_t}V - rV = 0 \hfill \cr V(T,S) = H(S) \hfill \cr} \right.

From equation (2), the total valuation adjustment U=V^V U = \hat V - V can be obtained, so the partial differential equation that U satisfies is: {tU+AtU(r+λB+λC)U=γXX+γKKλC(gCV)λB(gBV)r(X+ϕK+βX+βK)+λBεhU(T,S)=0 \left\{\matrix{{\partial _t}U + {A_t}U - \left({r + {\lambda _B} + {\lambda _C}} \right)U = - {\gamma _X}X + {\gamma _K}K - {\lambda _C}({g_C} - V) \hfill \cr - {\lambda _B}({g_B} - V) - r\left({X + \phi K + {\beta _X} + {\beta _K}} \right) + {\lambda _B}{\varepsilon _h} \hfill \cr U\left({T,S} \right) = 0 \hfill \cr} \right.

According to the Feynman-Kac formula, it can be found that the partial differential equation is expressed as Ut+μ(x,t)Ux+12σ2(x,t)2Ux2v(x,t)U=f(x,t) {{\partial U} \over {\partial t}} + \mu (x,t){{\partial U} \over {\partial x}} + {1 \over 2}{\sigma ^2}(x,t){{{\partial ^2}U} \over {\partial {x^2}}} - v(x,t)U = f(x,t)

When the boundary condition is U (x, T) = Ψ(x), the solution function of this partial differential equation can be written as the conditional expectation of a random process: U(x,t)=E[tTetsv(xt)dτf(xs,s)ds+etTv(xt)dτΨ(xT)|xt=x] U(x,t) = E\left[{\int_t^T {{e^{- \int_t^s {v\left({{x_t}} \right)d\tau}}}f\left({{x_s},s} \right)ds + {e^{- \int_t^T {v({x_t})d\tau}}}\Psi \left({{x_T}} \right)|{x_t} = x}} \right]

In the formula, μ, σ, υ, ψ and f (x, t) are known functions. T is the given parameter. x = (xt, t ≥ 0) is the Ito stochastic process determined by dXt = μ(Xt, t)dt +σ(Xt, t)dWt. The partial differential equation of the total valuation adjustment U is: Ut+AtU(r+λB+λC)U=γXX+γKKλC(gCV)λB(gBV)r(X+ϕK+βX+βK)+λBεh \eqalign{& {{\partial U} \over {\partial t}} + {A_t}U - \left({r + {\lambda _B} + {\lambda _C}} \right)U = \cr & - {\gamma _X}X + {\gamma _K}K - {\lambda _C}\left({{g_C} - V} \right) - {\lambda _B}\left({{g_B} - V} \right) - r\left({X + \phi K + {\beta _X} + {\beta _K}} \right) + {\lambda _B}{\varepsilon _h} \cr}

Boundary condition U (T, S) = 0. According to the Feynman-Kac formula: U(S,t)=E[tTetsv(S,τ)dτf(Sh,h)ds] U\left({S,t} \right) = E\left[{\int_t^T {{e^{- \int_t^s {v\left({S,\tau} \right)d\tau}}}f\left({{S_h},h} \right)ds}} \right]

Where: v(S,t)=r+λB+λCf(S,t)=γXX+γKKλC(gCV)λB(gBV)r(X+ϕK+βX+βK)+λBεhU(S,t)=E[ets(r+λB+λC)[γXX+γKKλC(gCV)r(X+ϕK+βX+βK)+λBεh]dσ] \eqalign{& v\left({S,t} \right) = r + {\lambda _B} + {\lambda _C} \cr & f\left({S,t} \right) = - {\gamma _X}X + {\gamma _K}K - {\lambda _C}\left({{g_C} - V} \right) - {\lambda _B}\left({{g_B} - V} \right) - r\left({X + \phi K + {\beta _X} + {\beta _K}} \right) + {\lambda _B}{\varepsilon _h} \cr & \,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,\,U\left({S,t} \right) = E\left[{{e^{- \int_t^s {\left({r + {\lambda _B} + {\lambda _C}} \right)}}}\left[\matrix{- {\gamma _X}X + {\gamma _K}K - {\lambda _C}\left({{g_C} - V} \right) - \hfill \cr r\left({X + \phi K + {\beta _X} + {\beta _K}} \right) + {\lambda _B}{\varepsilon _h} \hfill \cr} \right]d\sigma} \right] \cr}

Since U (S, t) is the total valuation adjustment, it can be decomposed into: U=CVA+FVA+MVA+COLVA+KVA U = CVA + FVA + MVA + COLVA + KVA

In CVA=Et[tTλC(u)ets(r+λB+λC)dτ(Vϕ(u)gC(u))du] CVA = - {E_t}\left[{\int_t^T {{\lambda _C}(u){e^{- \int_t^s {\left({r + {\lambda _B} + {\lambda _C}} \right)d\tau}}}\left({{V_\phi}(u) - {g_C}(u)} \right)du}} \right]

Credit valuation adjustment CVA is the present value of the expected expenses due to the default of the counterparty as estimated by the bank [13]. This reflects the impact of the counterparty's default of derivatives on the value of derivatives. The existence of counterparty credit risk in financial derivatives transactions will bring a certain degree of deviation to the pricing of derivatives contracts. In particular, the price of OTC derivatives trading contracts will deviate from its true value. Therefore, we need to make credit valuation adjustments to the value of derivative contracts. This makes its pricing more reasonable. FVA=Et[tT(rF(u)rϕ(u))ets(r+λB+λC)dτ(Vϕ(u)ϕ(u))du] FVA = - {E_t}\left[{\int_t^T {\left({{r_F}(u) - {r_\phi}(u)} \right){e^{- \int_t^s {\left({r + {\lambda _B} + {\lambda _C}} \right)d\tau}}}\left({{V_\phi}(u) - \phi (u)} \right)du}} \right]

Financing valuation adjustment FVA is the adjustment of the value of financing costs in derivatives transaction contracts [14]. This is a discount on expected financing costs. It reflects the impact of external financing costs on the value of derivatives. MVA=Et[tT(rF(u)rψ(u))ets(r+λB+λC)dτψB(u)du] MVA = - {E_t}\left[{\int_t^T {\left({{r_F}(u) - {r_\psi}(u)} \right){e^{- \int_t^s {\left({r + {\lambda _B} + {\lambda _C}} \right)d\tau}}}{\psi _B}(u)du}} \right]

Margin valuation adjustment MVA is the cost of issuing initial margin during the trading cycle. It reflects the financing cost of issuing initial margin for derivatives transactions. Therefore, it can also be called the FVA of the initial margin [15]. Especially when the transaction triggers additional initial reserves, margin valuation adjustments will occur. COLVA=Et[tT(γX(u)r(u))ets(r+λB+λC)dτX(u)+γX(u)βXdu] COLVA = - {E_t}\left[{\int_t^T {\left({{\gamma _X}(u) - r(u)} \right){e^{- \int_t^s {\left({r + {\lambda _B} + {\lambda _C}} \right)d\tau}}}X(u) + {\gamma _X}(u){\beta _X}du}} \right]

Collateral valuation adjustment COLVA is the cost incurred by the required collateral in the transaction cycle. KVA=Et[tT(γX(u)r(u))ets(r+λB+λC)dτK(u)γK(u)βK+λBεK(u)du] KVA = - {E_t}\left[{\int_t^T {\left({{\gamma _X}(u) - r(u)} \right){e^{- \int_t^s {\left({r + {\lambda _B} + {\lambda _C}} \right)d\tau}}}K(u) - {\gamma _K}(u){\beta _K} + {\lambda _B}{\varepsilon _K}(u)du}} \right]

The expression of capital valuation adjustment KVA is the discount value of capital, capital cash repurchase accounts, and capital adjustment εK (u). If a bank wants to conduct a transaction, it needs to hold a certain amount of equity capital during the transaction cycle to ensure the smooth progress of the transaction. This will involve the cost of holding equity capital. Banks need additional equity capital when investing in low-risk projects since bank equity holders require a certain return on equity investment. The return on these capitals must be at least higher than the return on equity investment required by the equity holders. The purpose of capital valuation adjustment is to ensure that the return on additional equity capital is not lower than the return on the basic equity investment [16]. The more accurate the calculation of capital valuation adjustments, the more conducive to rationalizing additional capital holdings.

Analysis of Numerical Examples

The price of bonds issued by fund companies depends on many factors. Mainly include asset volatility, bond maturity date, the weight of short-term debt and medium and long-term debt in the default boundary, the debt recovery rate at the time of default, market risk-free interest rate, financing leverage, etc. When drawing up capital structure plans, fund companies usually use debt ratios and debt financing ratios to assess bankruptcy risks. Assume that D1/S0=0.05 and D2/S0=0.1. We use Matlab software to program equation (16) and iteratively solve equation (17). When ρ = 0.15, 0.2, 0.25, 0.3, the influence of the simulated Hurst parameter H on the default probability of short-term bonds is shown in Figure 1 and Figure 2.

Figure 1

The impact of Hurst parameter H on short-term bond prices

Figure 2

The impact of Poisson jump parameter λ on short-term bond prices

It can be seen that the default probability of the short-term debt of the fund company gradually increases with the increase of the Hurst parameter H of the mixed fraction jump-diffusion fraction Brownian motion. When the risk factor becomes larger, the probability of default also rises. This is consistent with empirical evidence.

Conclusion

This article studies the cost of capital by including capital items in the hedging portfolio strategy. First, a hedging portfolio is designed from the perspective of an incomplete hedging strategy. The portfolio includes stocks, bonds with different recovery rates, stock options, capital, and different cash accounts. This article assumes that the stock price of the underlying asset obeys the jump-diffusion process to increase the rationality of asset price changes. This article also considers the risk of fluctuations in the underlying asset price. Based on this, stock options are added to the hedging portfolio to reduce the risk of stock price fluctuations.

Figure 1

The impact of Hurst parameter H on short-term bond prices
The impact of Hurst parameter H on short-term bond prices

Figure 2

The impact of Poisson jump parameter λ on short-term bond prices
The impact of Poisson jump parameter λ on short-term bond prices

Lu, Z., Yan, H., & Zhu, Y. European option pricing model based on uncertain fractional differential equation. Fuzzy Optimization and Decision Making., 2019; 18(2): 199–217 LuZ. YanH. ZhuY. European option pricing model based on uncertain fractional differential equation Fuzzy Optimization and Decision Making 2019 18 2 199 217 10.1007/s10700-018-9293-4 Search in Google Scholar

Constantinescu, C. D., Ramirez, J. M., & Zhu, W. R. An application of fractional differential equations to risk theory. Finance and Stochastics., 2019; 23(4): 1001–1024 ConstantinescuC. D. RamirezJ. M. ZhuW. R. An application of fractional differential equations to risk theory Finance and Stochastics 2019 23 4 1001 1024 10.1007/s00780-019-00400-8 Search in Google Scholar

Balcı, M. A. Fractional interaction of financial agents in a stock market network. Applied Mathematics and Nonlinear Sciences., 2020; 5(1): 317–336 BalcıM. A. Fractional interaction of financial agents in a stock market network Applied Mathematics and Nonlinear Sciences 2020 5 1 317 336 10.2478/amns.2020.1.00030 Search in Google Scholar

Gulian, M., Raissi, M., Perdikaris, P., & Karniadakis, G. Machine learning of space-fractional differential equations. SIAM Journal on Scientific Computing., 2019; 41(4): A2485–A2509 GulianM. RaissiM. PerdikarisP. KarniadakisG. Machine learning of space-fractional differential equations SIAM Journal on Scientific Computing 2019 41 4 A2485 A2509 10.1137/18M1204991 Search in Google Scholar

Ishii, H., & Roch, A. Existence and Uniqueness of Viscosity Solutions of an Integro-differential Equation Arising in Option Pricing. SIAM Journal on Financial Mathematics., 2021; 12(2): 604–640 IshiiH. RochA. Existence and Uniqueness of Viscosity Solutions of an Integro-differential Equation Arising in Option Pricing SIAM Journal on Financial Mathematics 2021 12 2 604 640 10.1137/20M1341441 Search in Google Scholar

Jalali, M. F. M., & Heidari, H. Predicting changes in Bitcoin price using grey system theory. Financial Innovation., 2020; 6(1): 1–12 JalaliM. F. M. HeidariH. Predicting changes in Bitcoin price using grey system theory Financial Innovation 2020 6 1 1 12 Search in Google Scholar

Zhang, Y., Gao, J., & Fu, Z. Valuing currency swap contracts in uncertain financial market. Fuzzy Optimization and Decision Making., 2019; 18(1): 15–35 ZhangY. GaoJ. FuZ. Valuing currency swap contracts in uncertain financial market Fuzzy Optimization and Decision Making 2019 18 1 15 35 10.1007/s10700-018-9284-5 Search in Google Scholar

Liu, X., Sun, H., Zhang, Y., & Fu, Z. A scale-dependent finite difference approximation for time fractional differential equation. Computational Mechanics., 2019; 63(3): 429–442 LiuX. SunH. ZhangY. FuZ. A scale-dependent finite difference approximation for time fractional differential equation Computational Mechanics 2019 63 3 429 442 10.1007/s00466-018-1601-x Search in Google Scholar

Yang, M., Zheng, Z., & Mookerjee, V. Prescribing response strategies to manage customer opinions: A stochastic differential equation approach. Information Systems Research., 2019; 30(2): 351–374 YangM. ZhengZ. MookerjeeV. Prescribing response strategies to manage customer opinions: A stochastic differential equation approach Information Systems Research 2019 30 2 351 374 10.1287/isre.2018.0805 Search in Google Scholar

Berner, J., Grohs, P., & Jentzen, A. Analysis of the Generalization Error: Empirical Risk Minimization over Deep Artificial Neural Networks Overcomes the Curse of Dimensionality in the Numerical Approximation of Black--Scholes Partial Differential Equations. SIAM Journal on Mathematics of Data Science., 2020; 2(3): 631–657 BernerJ. GrohsP. JentzenA. Analysis of the Generalization Error: Empirical Risk Minimization over Deep Artificial Neural Networks Overcomes the Curse of Dimensionality in the Numerical Approximation of Black--Scholes Partial Differential Equations SIAM Journal on Mathematics of Data Science 2020 2 3 631 657 10.1137/19M125649X Search in Google Scholar

Balter, A. G., & Pelsser, A. Pricing and hedging in incomplete markets with model uncertainty. European Journal of Operational Research., 2020; 282(3): 911–925 BalterA. G. PelsserA. Pricing and hedging in incomplete markets with model uncertainty European Journal of Operational Research 2020 282 3 911 925 10.1016/j.ejor.2019.09.054 Search in Google Scholar

Sun, Y., Shi, Y., & Zhang, Z. Finance big data: Management, analysis, and applications. International Journal of Electronic Commerce., 2019; 23(1): 9–11 SunY. ShiY. ZhangZ. Finance big data: Management, analysis, and applications International Journal of Electronic Commerce 2019 23 1 9 11 10.1080/10864415.2018.1512270 Search in Google Scholar

Liu, S., Borovykh, A., Grzelak, L. A., & Oosterlee, C. W. A neural network-based framework for financial model calibration. Journal of Mathematics in Industry., 2019; 9(1): 1–28 LiuS. BorovykhA. GrzelakL. A. OosterleeC. W. A neural network-based framework for financial model calibration Journal of Mathematics in Industry 2019 9 1 1 28 10.1186/s13362-019-0066-7 Search in Google Scholar

Shapiro, S. L., Reams, L., & So, K. K. F. Is it worth the price? The role of perceived financial risk, identification, and perceived value in purchasing pay-per-view broadcasts of combat sports. Sport Management Review., 2019; 22(2): 235–246 ShapiroS. L. ReamsL. SoK. K. F. Is it worth the price? The role of perceived financial risk, identification, and perceived value in purchasing pay-per-view broadcasts of combat sports Sport Management Review 2019 22 2 235 246 10.1016/j.smr.2018.03.002 Search in Google Scholar

Iglesias Martínez, M., Antonino-Daviu, J., de Córdoba, P. & Conejero, J. Higher-Order Spectral Analysis of Stray Flux Signals for Faults Detection in Induction Motors. Applied Mathematics and Nonlinear Sciences., 2020; 5(2): 1–14 Iglesias MartínezM. Antonino-DaviuJ. de CórdobaP. ConejeroJ. Higher-Order Spectral Analysis of Stray Flux Signals for Faults Detection in Induction Motors Applied Mathematics and Nonlinear Sciences 2020 5 2 1 14 10.2478/amns.2020.1.00032 Search in Google Scholar

Akkılıc, A., Sulaiman, T. & Bulut, H. Applications of the extended rational sine-cosine and sinh-cosh techniques to some nonlinear complex models arising in mathematical physics. Applied Mathematics and Nonlinear Sciences., 2021; 6(2): 19–30 AkkılıcA. SulaimanT. BulutH. Applications of the extended rational sine-cosine and sinh-cosh techniques to some nonlinear complex models arising in mathematical physics Applied Mathematics and Nonlinear Sciences 2021 6 2 19 30 10.2478/amns.2021.1.00021 Search in Google Scholar

Articoli consigliati da Trend MD

Pianifica la tua conferenza remota con Sciendo