Mafi Kyawun Zuba Jari ga Mai Inshora a Kasuwannin Kuɗi Biyu: Nazarin Sarrafa Stochastic

1. Gabatarwa

Wannan takarda tana magance gibi mai mahimmanci a cikin wallafe-wallafen sarrafa haɗarin inshora: dabarar zuba jari mafi kyau ga mai inshora da ke aiki a cikin kasuwannin kuɗi da yawa. Tsarin al'ada yawanci yana ƙuntata masu inshora zuwa kasuwannin cikin gida, yana yin watsi da rikitarwa da damfarin da ake samu daga hannun jarin ƙasashen waje. Marubutan sun faɗaɗa tsarin ragi na Cramér-Lundberg na gargajiya zuwa tsarin kusantarwa (diffusion approximation) kuma sun gabatar da yanayin kasuwar kuɗi biyu. Ƙimar musayar ƙasashen waje, wata hanya mai mahimmanci ta canzawa, an ƙirƙira ta da tafiyar stochastic (stochastic drift) wanda ke bin tsarin Ornstein-Uhlenbeck (OU), wanda ke ɗauke da halayen komawa ga matsakaici (mean-reverting) waɗanda ake yawan gani a kasuwannin kuɗi. Babban manufa ita ce haɓaka ƙimar amfanin ƙarshe na mai inshora (expected exponential utility of terminal wealth), wata ma'auni ta gama gari na gujewa haɗari a cikin mahallin kuɗi da na actuarial.

2. Tsarin Samfuri

2.1 Tsarin Ragi (Surplus Process)

Tsarin ragi na mai inshora $R(t)$ an ƙirƙira shi ta amfani da kusantarwa (diffusion approximation) na tsarin Cramér-Lundberg na gargajiya: $$dR(t) = c dt - d\left(\sum_{i=1}^{N(t)} Y_i\right) \approx \mu dt + \sigma_R dW_R(t)$$ inda $c$ shine ƙimar kuɗin inshora (premium rate), $\mu$ shine tafiya (drift), $\sigma_R$ shine saurin canji (volatility), kuma $W_R(t)$ shine motsi na Brownian na ma'auni. Wannan kusantarwa yana sauƙaƙa tsarin da'awar Poisson mai haɗaka (compound Poisson) don sauƙin nazari yayin da yake kiyaye mahimman halayen haɗari.

2.2 Kasuwar Kuɗi (Financial Market)

Mai inshora zai iya zuba jari a cikin:

1. Kayan Cikin Gida marasa Haɗari (Domestic Risk-Free Asset): $dB_d(t) = r_d B_d(t) dt$, tare da ƙimar riba mai tsayi $r_d$.
2. Kayan Ƙasashen Waje masu Haɗari (Foreign Risky Asset): $dS_f(t) = \mu_f S_f(t) dt + \sigma_f S_f(t) dW_f(t)$, inda $\mu_f$ da $\sigma_f$ suke tafiya (drift) da saurin canji (volatility), kuma $W_f(t)$ shine motsi na Brownian wanda ke da alaƙa da wasu matakai.

2.3 Tsarin Canjin Kuɗi (Exchange Rate Dynamics)

Wani sabon abu mai mahimmanci shine ƙirƙirar ƙimar musayar kuɗi $X(t)$ (raka'a na kuɗin cikin gida a kowace raka'a na kuɗin ƙasashen waje). Tsarinta shine: $$dX(t) = X(t)[\theta(t) dt + \sigma_X dW_X(t)]$$ inda matsakaicin ƙimar girma na lokaci-lokaci $\theta(t)$ shi kansa yana bin tsarin Ornstein-Uhlenbeck: $$d\theta(t) = \kappa(\bar{\theta} - \theta(t)) dt + \sigma_\theta dW_\theta(t)$$ A nan, $\kappa>0$ shine saurin komawa ga matsakaici (mean-reversion speed), $\bar{\theta}$ shine matsakaici na dogon lokaci (long-term mean), kuma $\sigma_\theta$ shine saurin canji (volatility). Motsin Brownian $W_R, W_f, W_X, W_\theta$ suna da alaƙa, suna gabatar da haɗin kai tsakanin haɗarin inshora, haɗarin kasuwa, da haɗarin kuɗi.

3. Matsalar Haɓakawa (Optimization Problem)

3.1 Aikin Manufa (Objective Function)

Bari $W(t)$ ya zama jimlar dukiyar mai inshora a cikin kuɗin cikin gida a lokacin $t$, kuma $\pi(t)$ ya zama adadin da aka zuba a cikin kayan ƙasashen waje masu haɗari. Manufar ita ce nemo dabarar da ta fi dacewa $\pi^*(t)$ wacce ke haɓaka ƙimar amfanin ƙarshe na dukiya a lokacin $T$: $$\sup_{\pi(t)} \mathbb{E}\left[ -\frac{1}{\gamma} e^{-\gamma W(T)} \right]$$ inda $\gamma > 0$ shine ma'aunin ƙin haɗari na cikakke mai tsayi (constant absolute risk aversion coefficient). An zaɓi aikin amfani mai ƙima (exponential utility function) saboda sauƙin nazarinsa da kuma halinsa na cikakken ƙin haɗari mai tsayi.

3.2 Daidaitaccen Hamilton-Jacobi-Bellman (HJB Equation)

Ta amfani da ƙa'idar tsara tsari mai ƙarfi (dynamic programming principle), an ayyana aikin ƙima $V(t, w, \theta)$ a matsayin mafi girman ƙimar amfani da ake tsammani daga lokacin $t$ tare da dukiya $w$ da mai canza yanayi $\theta$. Daidaitaccen HJB shine: $$\sup_{\pi} \left\{ V_t + \mathcal{L} V \right\} = 0$$ tare da sharadi na ƙarshe $V(T, w, \theta) = -\frac{1}{\gamma}e^{-\gamma w}$. A nan, $\mathcal{L}$ shine mai haifar da ƙaramin tsari (infinitesimal generator) na tsarin dukiya $W(t)$ da tsarin OU $\theta(t)$, wanda ya haɗa duk tafiyoyi (drift), watsawa (diffusion), da sharuɗɗan alaƙa daga tsarin samfurin.

4. Maganin Nazari (Analytical Solution)

4.1 Dabarar Zuba Jari Mafi Kyau (Optimal Investment Strategy)

Ta hanyar warware daidaitaccen HJB, an samo zuba jari mafi kyau a cikin kayan ƙasashen waje masu haɗari kamar haka: $$\pi^*(t) = \frac{1}{\gamma \sigma_f^2} \left[ \mu_f - r_d + \rho_{Xf}\sigma_X\sigma_f + \frac{\Psi(t)}{\gamma} \right]$$ inda $\rho_{Xf}$ shine alaƙar tsakanin ƙimar musayar kuɗi da kayan ƙasashen waje, kuma $\Psi(t)$ wani aiki ne wanda ya taso daga tafiyar stochastic $\theta(t)$ na ƙimar musayar kuɗi, yawanci yana ɗauke da sharuɗɗa kamar $\frac{\partial V/\partial \theta}{\partial V/\partial w}$. Wannan dabarar tana da tsarin matsakaici-bambanci na gargajiya (kalmarta ta farko) wanda aka daidaita don haɗarin canjin kuɗi (kalmarta ta biyu) da kuma wani ɓangare na kariya mai ƙarfi (dynamic hedging component) don tafiyar stochastic (kalmarta ta uku).

4.2 Aikin Ƙima (Value Function)

Aikin ƙima ya yarda da siffar exponential affine: $$V(t, w, \theta) = -\frac{1}{\gamma} \exp\left\{ -\gamma w e^{r_d (T-t)} + A(t) \theta + B(t) \theta^2 + C(t) \right\}$$ inda $A(t)$, $B(t)$, da $C(t)$ ayyuka ne na ƙayyadaddun lokaci (deterministic functions of time) waɗanda ke gamsar da tsarin daidaitattun bambance-bambancen Riccati (Riccati ODEs) waɗanda aka samo daga daidaitaccen HJB. Maganganunsu suna siffanta tasirin tafiyar stochastic na ƙimar musayar kuɗi akan mafi girman ƙimar amfani da ake tsammani na mai inshora.

5. Nazarin Lambobi (Numerical Analysis)

Takardar ta gabatar da nazarin lambobi don kwatanta halayen dabarar da ta fi dacewa. Maɓuɓɓukan ma'auni na iya zama kamar: $\gamma=2$, $r_d=0.03$, $\mu_f=0.08$, $\sigma_f=0.2$, $\sigma_X=0.15$, $\kappa=0.5$, $\bar{\theta}=0.01$, $\sigma_\theta=0.05$, $T=5$. Nazarin yana iya nuna:

• Hankali ga Ƙin Haɗari ($\gamma$): Yayin da $\gamma$ ya ƙaru (mafi ƙin haɗari), $\pi^*$ yana raguwa.
• Tasirin Komawa ga Matsakaici ($\kappa$): Mafi girma $\kappa$ (saurin komawa ga matsakaici) yana rage rashin tabbas daga $\theta(t)$, yana iya ƙara rabon zuba jari.
• Tasirin Alaƙa ($\rho_{Xf}$): Alaƙa mai kyau tsakanin kayan ƙasashen waje da ƙimar musayar kuɗi tana ƙara haɗarin aiki, yana iya rage $\pi^*$.
• Tsarin Dukiya (Wealth Dynamics): Hanyoyin da aka ƙirƙira na dukiya $W(t)$ a ƙarƙashin dabarar da ta fi dacewa da kuma dabarar butulci (misali, rabo mai tsayi) za su nuna mafi kyawun rarraba dukiyar ƙarshe da ƙananan haɗarin faɗuwa mai tsanani ga dabarar da ta fi dacewa.

Duk da cewa ba a ba da takamaiman adadi a cikin ɓangaren da aka cire ba, irin wannan nazari yawanci ya ƙunshi zana $\pi^*(t)$ akan lokaci don saiti daban-daban na ma'auni da kuma kwatanta rarraba $W(T)$ ta hanyar simintin Monte Carlo.

6. Fahimta ta Asali & Ra'ayin Manazarcin

Fahimta ta Asali: Wannan takarda ba wani ƙarin gyara ne kawai ga samfurin zuba jari na mai inshora ba. Ƙimarta ta asali tana cikin haɗa haɗarin kuɗi na stochastic a cikin tsarin sarrafa kadarori da alawus na mai inshora (ALM). Ta hanyar ƙirƙirar tafiyar ƙimar musayar kuɗi a matsayin tsarin OU, sun wuce ra'ayoyin motsi na Brownian na geometric ko na tsayi, suna ɗaukar gaskiya ta asali ga masu inshora na duniya: yanayin kuɗi yana dawwama amma ba shi da tabbas, yana tasiri daga abubuwan tattalin arziƙi kamar bambance-bambancen hauhawar farashin kayayyaki da ma'auni na biyan kuɗi—abubuwan da aka ambata a fili a cikin gabatarwa. Dabarar da aka samo ba dabara ce ta tsaye ba; tacewa ce mai ƙarfi wacce ke daidaitawa akai-akai bisa ga kimantaccen yanayi $\theta(t)$ na ƙarfin kasuwar kuɗi.

Kwararar Ma'ana: Gine-ginen samfurin yana da ma'ana. Ya fara da tushe mai ƙarfi (kusantarwa-diffusion na Cramér-Lundberg) kuma yana sanya rikitarwa cikin tsari: da farko gabatar da kayan ƙasashen waje, sannan ƙimar musayar kuɗi na stochastic, kuma a ƙarshe ya sa tafiyar ƙimar musayar kuɗi ta zama stochastic. Wannan tsarin mataki-mataki yana sa daidaitaccen HJB ya zama mai sauƙi kuma maganin ƙarshe ya zama mai fassara. Maganin ya rarraba yanke shawarar zuba jari zuwa kalmar matsakaici-bambanci mai hangen nesa (myopic mean-variance term), kariya ta tsaye don haɗarin kuɗi na lokaci-lokaci, da kuma kariya mai ƙarfi don tafiyar da ke tasowa—wani rarrabuwa mai kama da matsalar fayil na Merton tare da damfarin zuba jari na stochastic.

Ƙarfafawa & Kurakurai: Ƙarfafawa: 1) Gaskiya: Tsarin OU don $\theta(t)$ babban haɓaka ne a cikin ƙirƙirar gaskiya fiye da samfuran tafiya mai tsayi. 2) Sauƙi: Cimma magani mai rufaffiyar fomula (closed-form solution) don irin wannan matsala mai yawa na sarrafa stochastic (multi-factor stochastic control problem) babban nasara ce ta fasaha. 3) Sakamako mai Aiki: Dabarar da ta fi dacewa $\pi^*(t)$ an bayyana ta cikin sharuɗɗan ma'auni masu iya gani ko kimantawa. Kurakurai & Iyakoki: 1) Kusantarwa (Diffusion Approximation): Haɗarin inshora na asali an sauƙaƙa shi zuwa watsawa (diffusion), yana shafe haɗarin tsalle (manyan da'awar) wanda ke tsakiyar inshora. Wannan babban rangwame ne don sauƙi, kamar yadda aka yarda a cikin ayyuka kamar Schmidli (2008). 2) Kayan Haɗari Guda ɗaya: Samfurin yana la'akari da kayan ƙasashen waje masu haɗari guda ɗaya kawai. Faɗaɗa zuwa kaya da yawa (multi-asset extension), ko da yake yana da nauyi a lissafi, yana da mahimmanci don gina fayil na aiki. 3) Hankalin Ma'auni: Dabarar ta dogara sosai akan kimanta ma'auni na tsarin OU ($\kappa, \bar{\theta}, \sigma_\theta$) daidai, wanda ke da wahala a cikin kuɗi. Kuskuren kimantawa zai iya haifar da yanke shawara mara kyau sosai. 4) Babu Gurbacewa: Yana yin watsi da farashin ciniki da haraji, waɗanda ke da mahimmanci a cikin zuba jari na ketare iyaka.

Fahimta mai Aiki: Ga Babban Jami'in Zuba Jari a wani mai inshora na duniya, wannan bincike yana ba da tsarin ƙididdiga na ƙididdiga don tabbatar da shirye-shiryen kariya na kuɗi mai ƙarfi. Samfurin yana nuna cewa bai kamata kariya ta kasance ta tsaye ba amma ya kamata ta bambanta da "ƙarfin tsiya" da ake ganin kasuwar kuɗi tana da shi. A aikace, wannan yana nufin:

1. Kafa tsarin ƙididdiga don kimanta mai canza yanayi $\theta(t)$ (tafiyar kuɗi) ta amfani da samfuran tattalin arziƙi akan bayanan lokaci-lokaci.
2. Yi amfani da dabara da aka samo don lissafta rabon kariya mai ƙarfi, ba na tsaye ba.
3. Gane iyakar samfurin game da haɗarin tsalle. Haɗa wannan dabarar tare da kayan aikin kariya na haɗarin wutsiya (misali, zaɓuɓɓuka) don karewa daga manyan abubuwan da'awar da kusantarwa (diffusion approximation) ya rasa.
4. Yi la'akari da sakamakon samfurin a matsayin ma'auni na dabarar maimakon siginar ciniki na zahiri, daidaitawa don gurbacewar duniyar gaske da ƙuntatawa na kaya da yawa.

Takardar tana canza tattaunawar daga "ko" don kare haɗarin kuɗi zuwa "yadda ake kare shi da ƙarfi" bisa ga ra'ayi na stochastic na tushen kuɗi.

7. Cikakkun Bayanai na Fasaha & Ƙirƙirar Lissafi

Tsarin dukiya $W(t)$ a cikin kuɗin cikin gida, bayar da dabarar $\pi(t)$, yana tasowa kamar haka: $$dW(t) = [W(t) - \pi(t)] r_d dt + \pi(t) \frac{dS_f(t)}{S_f(t)} + \pi(t) \frac{dX(t)}{X(t)} + dR(t)$$ Maye gurbin tsarin $S_f(t)$, $X(t)$, da $R(t)$, da kuma amfani da ƙa'idar Itô don matakai masu alaƙa, yana ba da cikakken SDE don $W(t)$. Daidaitaccen HJB ya zama: $$ \begin{aligned} 0 = \sup_{\pi} \Bigg\{ & V_t + \left[ w r_d + \pi(\mu_f - r_d + \theta) + \mu \right] V_w + \kappa(\bar{\theta} - \theta) V_\theta \\ & + \frac{1}{2}\left[ \pi^2(\sigma_f^2 + \sigma_X^2 + 2\rho_{fX}\sigma_f\sigma_X) + \sigma_R^2 + 2\pi(\rho_{fR}\sigma_f\sigma_R + \rho_{XR}\sigma_X\sigma_R) \right] V_{ww} \\ & + \frac{1}{2}\sigma_\theta^2 V_{\theta\theta} + \pi \sigma_\theta (\rho_{f\theta}\sigma_f + \rho_{X\theta}\sigma_X) V_{w\theta} \Bigg\} \end{aligned} $$ Sharaɗin mataki na farko don mafi girma yana haifar da magana don $\pi^*(t)$ da aka nuna a Sashe na 4.1. Maye gurbin $\pi^*$ a cikin daidaitaccen HJB kuma ɗaukar siffar exponential affine don $V$ yana kaiwa ga daidaitattun Riccati ODEs don $A(t)$, $B(t)$, $C(t)$.

8. Tsarin Nazari: Wani Lamari na Aiki

Yanayi: Wani mai inshora na Japan wanda ba na rayuwa ba (kuɗin cikin gida: JPY) yana riƙe da ajiyar jari kuma yana son zuba wani ɓangare a cikin kasuwar hannun jari ta Amurka (fihirisar S&P 500, wanda aka ƙidaya a USD) don neman riba mafi girma. Mai inshora yana fuskantar da'awar alawus a JPY.

Aiwatar da Tsarin:

1. Daidaituwar Samfuri (Model Calibration):
   • Ragi (Surplus R): Kimanta $\mu$ da $\sigma_R$ daga bayanan riba/asara na rubutu na tarihi.
   • Kayan Ƙasashen Waje (S&P 500): Yi amfani da dawowar tarihi don kimanta $\mu_f$ da $\sigma_f$.
   • Ƙimar Musayar Kuɗi (USD/JPY): Ƙirƙiri $X(t)$. Kimanta $\sigma_X$ daga saurin canjin FX (FX volatility). Mahimmanci shine kimanta tsarin OU don $\theta(t)$. Ana iya yin haka ta amfani da tacewa Kalman ko mafi girman kimantawar ƙima akan dawowar FX na tarihi, ana ɗaukar $\theta(t)$ a matsayin mai canza yanayi na ɓoye wanda ke motsa shi ta abubuwan tattalin arziƙi kamar bambance-bambancen hauhawar farashin kayayyaki na Amurka-Japan da ma'auni na ciniki.
   • Alaƙa: Kimanta alaƙa ($\rho_{fX}, \rho_{fR},$ da sauransu) daga bayanan tarihi tsakanin dawowar S&P 500, canje-canjen USD/JPY, da lokutan biyan da'awar mai inshora.
   • Ƙin Haɗari ($\gamma$): Saita bisa ga juriyar haɗari na kamfani, wanda za a iya samo shi daga matsayin ƙimar kiredit da aka yi niyya ko ma'aunin isasshen jari.
2. Aiwatar da Dabarar: A kowane lokacin yanke shawara (misali, na kwata):
   1. Sabunta kimantaccen yanayi na ɓoye $\theta(t)$ (tafiyar USD na yanzu).
   2. Saka duk ma'auni da aka daidaita da $\theta(t)$ na yanzu a cikin dabara don $\pi^*(t)$.
   3. Daidaituwar bayyanar hannun jari na USD don dacewa da $\pi^*(t)$ a matsayin kashi na dukiyar yanzu $W(t)$.
3. Sa ido kan Aiki (Performance Monitoring): Gwada dabarar a baya akan ma'auni na tsaye (misali, rabon 20% na tsaye zuwa hannun jari na USD) a kan lokutan tarihi waɗanda ke da tsarin FX daban-daban (misali, ƙarfafa USD da raunana USD). Maɓuɓɓukan ma'auni: Amfanin dukiyar ƙarshe, ma'aunin Sharpe, matsakaicin faɗuwa, da Ƙimar-a-Haɗari (Value-at-Risk).

Wannan lamari yana nuna yadda samfurin na ka'ida ke fassara shi zuwa tsarin zuba jari mai tsari, mai maimaitawa ga mai inshora na duniya.

9. Aikace-aikacen Gaba & Jagororin Bincike

Aikace-aikace:

• ALM mai Ƙarfi ga Masu Inshora na Ƙasashen Duniya: Aikace-aikacen kai tsaye ga masu inshora masu alawus da zuba jari a cikin kuɗi da yawa.
• Asusun Fansho tare da Bayyanar Duniya: Irin wannan matsalolin zuba jari da ke haifar da alawus suna wanzuwa ga asusun fansho waɗanda ke amfanar membobi a ƙasashe daban-daban.
• Ƙirƙirar Kayayyakin Zuba Jari masu Kariya ta Kuɗi: Masu sarrafa kadarori za su iya amfani da ma'anar samfurin don ƙirƙirar hannun jari na ƙasashen waje ko na bond da aka kare da ƙarfi wanda aka keɓance don abokan ciniki na hukumomi kamar masu inshora.
• Ƙirƙirar Ma'aunin Jari na Tsari: Tsarin zai iya ba da labari game da ƙirƙirar samfuran ciki don lissafta Bukatar Jari na Solvency (SCR) don haɗarin kasuwa a ƙarƙashin tsarin kamar Solvency II, musamman don ɓangaren haɗarin kuɗin ƙasashen waje.

Jagororin Bincike:

1. Haɗa Haɗarin Tsalle: Mafi mahimmanci faɗaɗa shine sake haɗa matakai na tsalle don da'awar inshora, matsawa daga kusantarwa (diffusion approximation) zuwa samfuri kamar Poisson mai haɗaka ko tsarin Lévy. Wannan zai dace da ayyukan kwanan nan a cikin lissafin inshora, kamar waɗanda aka taƙaita a cikin Asmussen & Albrecher (2010).
2. Kaya masu Haɗari da Kuɗi da Yawa: Faɗaɗa samfurin zuwa kwandon kayan ƙasashen waje a cikin kuɗi daban-daban zai ƙara dacewa da aiki.
3. Koyo da Rashin Tabbaci: Haɗa rashin tabbas na ma'auni da koyo, inda mai inshora bai san ainihin ma'auni na tsarin OU ba amma yana sabunta imaninsa akan lokaci (koyon Bayesian), kamar yadda aka bincika a cikin samfuran ƙin rashin tabbas.
4. Ƙirƙirar Masu Tuka Tattalin Arziƙi a fili: Maimakon tsarin OU na ɓoye don $\theta(t)$, a ƙirƙira shi kai tsaye a matsayin aiki na masu canjin tattalin arziƙi masu iya gani (bambance-bambancen hauhawar farashin kayayyaki, faɗaɗa ƙimar riba, ma'auni na ciniki), ƙirƙirar samfuri mafi mahimmanci kuma mai gwadawa.
5. Haɗa Gurbacewa: Ƙara farashin ciniki na rabo ko haraji zai sa dabarar ta zama mafi gaskiya, yana iya haifar da yankin "babu ciniki" a kusa da rabon da ya fi dacewa.

10. Nassoshi

1. Zhou, Q., & Guo, J. (2020). Optimal Control of Investment for an Insurer in Two Currency Markets. arXiv preprint arXiv:2006.02857.
2. Browne, S. (1995). Optimal investment policies for a firm with a random risk process: exponential utility and minimizing the probability of ruin. Mathematics of Operations Research, 20(4), 937-958.
3. Schmidli, H. (2008). Stochastic Control in Insurance. Springer Science & Business Media.
4. Merton, R. C. (1971). Optimum consumption and portfolio rules in a continuous-time model. Journal of Economic Theory, 3(4), 373-413.
5. Asmussen, S., & Albrecher, H. (2010). Ruin Probabilities (2nd ed.). World Scientific.
6. Yang, H., & Zhang, L. (2005). Optimal investment for insurer with jump-diffusion risk process. Insurance: Mathematics and Economics, 37(3), 615-634.
7. Bai, L., & Guo, J. (2008). Optimal proportional reinsurance and investment with multiple risky assets and no-shorting constraint. Insurance: Mathematics and Economics, 42(3), 968-975.
8. Hipp, C., & Plum, M. (2000). Optimal investment for insurers. Insurance: Mathematics and Economics, 27(2), 215-228.