本文關(guān)鍵詞：資源分配復(fù)雜系統(tǒng)的統(tǒng)計(jì)性質(zhì)研究　出處：《復(fù)旦大學(xué)》2012年博士論文　論文類型：學(xué)位論文

  更多相關(guān)文章： 經(jīng)濟(jì)物理 基于主體的建模 行為人模擬 市場(chǎng)導(dǎo)向的資源分配博弈 資源分配 可控實(shí)驗(yàn) 異質(zhì)偏好 經(jīng)濟(jì)實(shí)驗(yàn) 看不見的手

【摘要】：經(jīng)濟(jì)物理學(xué)是一門新興的交叉學(xué)科,致力于用物理學(xué)的世界觀和方法論來重新審視那些尚未妥善解決的經(jīng)濟(jì)學(xué)問題。物理學(xué)家對(duì)經(jīng)濟(jì)學(xué)的興趣,源自于物理學(xué)家對(duì)復(fù)雜系統(tǒng)的興趣。事實(shí)上,很多社會(huì)、經(jīng)濟(jì)、以及生物系統(tǒng)都包含著大量相互作用的個(gè)體(也叫主體,agent),這些系統(tǒng)都屬于復(fù)雜自適應(yīng)系統(tǒng)(complex adaptive system, CAS)。一個(gè)主體的行為往往既受到系統(tǒng)內(nèi)其他主體行為的影響,又受到外界環(huán)境的影響。復(fù)雜自適應(yīng)系統(tǒng)在宏觀上具有豐富的統(tǒng)計(jì)性質(zhì),而在微觀上則具有十分有趣的動(dòng)力學(xué)機(jī)制。自誕生至今的大約20年里,經(jīng)濟(jì)物理學(xué)形成了一些對(duì)由我們?nèi)祟愖陨斫M成的復(fù)雜系統(tǒng)行之有效的研究方法：對(duì)經(jīng)濟(jì)數(shù)據(jù)的統(tǒng)計(jì)分析；基于主體的建模(agent-based modeling,也可叫做行為人模擬)；可控實(shí)驗(yàn)(controlled experiments)。這些方法與傳統(tǒng)的經(jīng)濟(jì)學(xué)研究方法非常不同,但是卻有希望幫助我們更好地理解經(jīng)濟(jì)系統(tǒng)。 本文中,我們主要研究了中國(guó)股票市場(chǎng)的一些統(tǒng)計(jì)性質(zhì),以及資源分配問題背后隱含的“看不見的手”的微觀機(jī)制,并提出了市場(chǎng)導(dǎo)向的資源分配博弈(market-directed resource allocation game, MDRAG)模型。本文按照我在經(jīng)濟(jì)物理領(lǐng)域開展研究工作的順序進(jìn)行章節(jié)安排。在我們的工作中,上述經(jīng)濟(jì)物理學(xué)領(lǐng)域的三種方法都有涉及。 第一章,作為緒論,主要回答了三個(gè)問題：經(jīng)濟(jì)物理學(xué)是做什么的,經(jīng)濟(jì)學(xué)為什么需要物理學(xué)家,以及物理學(xué)家應(yīng)該如何研究經(jīng)濟(jì)學(xué)。 第二章,主要介紹我們?cè)谕瓿蒑DRAG前的早期工作：首先是回顧了作為基于主體模型的代表作“少數(shù)派博弈”(minority game, MG)模型及其相變行為；然后將其他科學(xué)工作者對(duì)西方成熟股票市場(chǎng)的統(tǒng)計(jì)經(jīng)驗(yàn)運(yùn)用到中國(guó)股票市場(chǎng),對(duì)中國(guó)股票市場(chǎng)在牛市和熊市中股指收益率的統(tǒng)計(jì)規(guī)律進(jìn)行分析比較；最后從原始的MG模型出發(fā)嘗試模擬真實(shí)股票市場(chǎng),并檢驗(yàn)?zāi)Ｐ彤a(chǎn)生的收益率序列是否具有真實(shí)股價(jià)應(yīng)有的特征事實(shí)(stylized facts) 第三章,詳細(xì)介紹了關(guān)于MDRAG的工作。在這項(xiàng)工作中,我們聚焦于資源分配的問題。首先,我們?cè)O(shè)計(jì)并組織了一系列行為經(jīng)濟(jì)學(xué)的真人實(shí)驗(yàn),找到了“看不見的手”存在的證據(jù)。然后,我們?cè)贛G模型的基礎(chǔ)上構(gòu)建了MDRAG模型。新模型很好地解釋了真人實(shí)驗(yàn)結(jié)果。與此同時(shí),通過MDRAG模擬,我們提出了“看不見的手”導(dǎo)向的一種可能的微觀機(jī)制,并且發(fā)現(xiàn)了使其發(fā)揮調(diào)節(jié)功效的充分條件：第一(也是最重要的)：資源分配市場(chǎng)的參與者必須具備對(duì)各種資源的異質(zhì)偏好(heterogeneous preference)及根據(jù)環(huán)境演化改變偏好的能力；第二：市場(chǎng)參與者必須具備與環(huán)境的復(fù)雜程度相匹配的決策能力。更為有趣的是,在MDRAG的模擬結(jié)果中我們發(fā)現(xiàn)了一系列的相變,正是在這些相變的臨界點(diǎn)附近“看不見的手”的調(diào)節(jié)功效可以被發(fā)揮到極致,此時(shí)市場(chǎng)呈現(xiàn)出有效、穩(wěn)定、不可預(yù)測(cè)的狀態(tài)。 第四章,簡(jiǎn)單介紹了基于MDRAG的一些后續(xù)工作,例如對(duì)跟風(fēng)行為的研究等。 在最后的第五章,我們對(duì)經(jīng)濟(jì)物理學(xué)的發(fā)展以及我們研究中積累的經(jīng)驗(yàn)進(jìn)行一些回顧,并對(duì)領(lǐng)域內(nèi)某些方面在未來的發(fā)展提出一些展望。
[Abstract]:In fact , many social , economic , and biological systems contain a large number of interacting individuals ( also called subjects , agents ) , all of which belong to complex adaptive systems ( CAS ) . The behavior of a main body is often influenced by other subjects in the system , but also influenced by the external environment . The complex adaptive system has abundant statistical properties on the macro level , and has a very interesting dynamic mechanism on the microscopic scale . Since the birth to the present , economic physics has formed some research methods which are effective for the complex system composed of our human beings : statistical analysis of economic data ; agent - based modeling , also called actor simulation ; controlled experiments . These approaches are very different from traditional economics research methods , but there is a desire to help us understand the economic system better . In this paper , we mainly study the statistical nature of the Chinese stock market and the hidden invisible hand behind the resource allocation problem , and put forward the market - directed resource allocation game ( MDRAG ) model . Chapter one , as an introduction , mainly answers three questions : what economics physics is doing , why economics needs physicists , and how physicists should study economics . In chapter 2 , we mainly introduce our early work before completing MDRAG : Firstly , we review the minority game ( MG ) model and its phase change behavior as the representative of the main body model , then apply the statistical experience of other scientific workers to the mature stock market of the West to analyze and compare the statistical laws of stock index yield in bull market and bear market . Finally , we try to simulate the real stock market from the original MG model , and verify whether the yield sequence generated by the model has the characteristic facts that the real stock price should have . Chapter three introduces the work of MDRAG . In this work , we focus on the problem of resource allocation . First , we design and organize a series of real - person experiments of behavioral economics , find out the evidence of " invisible hand " . At the same time , we have developed MDRAG model on the basis of MG model . In chapter 4 , some follow - up tasks based on MDRAG are briefly introduced , such as the research on the behavior of high - heeled wind . In the last chapter , we review the development of economic physics and the experience accumulated in our research , and put forward some prospects for the future development of some aspects in the field .

【學(xué)位授予單位】：復(fù)旦大學(xué)
【學(xué)位級(jí)別】：博士
【學(xué)位授予年份】：2012
【分類號(hào)】：F832.51;F224
，

本文編號(hào)：1419207