Submitted by Fernando Reis de Odriozola (odriozola.fernando@gmail.com) on 2015-09-21T04:39:27Z No. of bitstreams: 1 Dissertação - Fernando R Odriozola.pdf: 881210 bytes, checksum: 13c5e46a6da326c976883920a7af7eb6 (MD5) / Approved for entry into archive by Renata de Souza Nascimento (renata.souza@fgv.br) on 2015-09-21T23:06:54Z (GMT) No. of bitstreams: 1 Dissertação - Fernando R Odriozola.pdf: 881210 bytes, checksum: 13c5e46a6da326c976883920a7af7eb6 (MD5) / Made available in DSpace on 2015-09-22T13:32:47Z (GMT). No. of bitstreams: 1 Dissertação - Fernando R Odriozola.pdf: 881210 bytes, checksum: 13c5e46a6da326c976883920a7af7eb6 (MD5) Previous issue date: 2015-08-24 / For complex systems, traditional analytical-approach with differential equations sometimes results in intractable solutions. An alternative approach could be through Agents-Based Models as a complementary tool witch systems can be modeled from their constituent parts and interactions. Financial Markets are good examples of complex system and thus Agent-Based Models would be a correct approach. This paper implements an Artificial Financial Market composed by market makers, information broadcasters and a set of heterogeneous agents who trade assets through a Continuous Double Auction mechanism. Several aspects of the simulation were investigated to consolidate their understanding and thus contribute to the design of models, where we can highlight, among others: distinctions between Discrete and Continuous Double Auction; implications of Market Maker spread settings; Budget Constraints effects on agents and Analysis of pricing formation in offer submissions. Thinking about the adherence of the model to the Brazilian market reality, a method named Inverse Simulation is used to calibrate the input parameters in a way that the output matches historical market price series. / Na modelagem de sistemas complexos, abordagens analíticas tradicionais com equações diferenciais muitas vezes resultam em soluções intratáveis. Para contornar este problema, Modelos Baseados em Agentes surgem como uma ferramenta complementar, onde o sistema é modelado a partir de suas entidades constituintes e interações. Mercados Financeiros são exemplos de sistemas complexos, e como tais, o uso de modelos baseados em agentes é aplicável. Este trabalho implementa um Mercado Financeiro Artificial composto por formadores de mercado, difusores de informações e um conjunto de agentes heterogêneos que negociam um ativo através de um mecanismo de Leilão Duplo Contínuo. Diversos aspectos da simulação são investigados para consolidar sua compreensão e assim contribuir com a concepção de modelos, onde podemos destacar entre outros: Diferenças do Leilão Duplo Contínuo contra o Discreto; Implicações da variação do spread praticado pelo Formador de Mercado; Efeito de Restrições Orçamentárias sobre os agentes e Análise da formação de preços na emissão de ofertas. Pensando na aderência do modelo com a realidade do mercado brasileiro, uma técnica auxiliar chamada Simulação Inversa, é utilizada para calibrar os parâmetros de entrada, de forma que trajetórias de preços simulados resultantes sejam próximas à séries de preços históricos observadas no mercado.

「預測市場」以未來事件為交易標的，透過網路平台彙整即時資訊，運用價格來判斷未來事件的發展，此研究方法同時具有「適當獎懲」與「連續修正」兩項特性。 本研究以文獻分析途徑探究預測市場在不同交易機制下的運作方式與市場價格決定過程，並據此分析不同交易機制之間的差異與影響預測市場運作的因素；除此之外，本研究另以個案研究途徑來探討「連續雙向拍賣」與「市場計分法則」兩個交易機制在價格準確度、市場流動性、價格炒作、參與誘因與莊家風險之間的差異。 本研究發現：若交易機制是連續雙向拍賣，則「0-100型」合約價格的預測準確度較高；若交易機制是市場計分法則，則「落點預測型」合約價格的預測準確度較高。連續雙向拍賣機制具有市場流動性不足的問題；市場計分法則機制面臨莊家風險的危機且不適用於市場競爭度高的環境；而上述兩種交易機制皆會出現價格炒作的現象。 / “Prediction market” is a research method based on immediate information collecting and organizing on the internet platform. With future events as the object of transaction, variations of the price of each transaction thus immediately provide the prediction of the development of future events. Therefore, this method has two properties including “appropriate incentives” and “continuous correction”. In this study, document analysis is first conducted to review the operation modes of different trading mechanisms for prediction markets and the process of price making. Accordingly, differences between trading mechanisms and the factors that affect the operation of prediction market will also be analyzed. Furthermore, comparisons of the price accuracy, market liquidity, price speculation, incentives and maker risks between "continuous double auction" and "market scoring rule" are discussed in case study. The findings of this study: if the trading mechanism is “continuous double auction”, the price accuracy of “winner-take-all” contract is higher; if the trading mechanism is “market scoring rules”, the price accuracy of “index” contract is higher. There exists insufficient market liquidity in “continuous double auction;” while in “market scoring rule,” there exists maker risks and it is hard to be applied in highly competitive market. The phenomenon of price speculation exits in both trading mechanisms.

Submitted by Milton Saito (milton.saito@gvmail.br) on 2013-09-11T03:21:58Z No. of bitstreams: 1 dissertacao-MiltonSaito.pdf: 8406128 bytes, checksum: 2c7b006b9cca9d097d0ee1116e898ef2 (MD5) / Approved for entry into archive by Suzinei Teles Garcia Garcia (suzinei.garcia@fgv.br) on 2013-09-11T13:27:29Z (GMT) No. of bitstreams: 1 dissertacao-MiltonSaito.pdf: 8406128 bytes, checksum: 2c7b006b9cca9d097d0ee1116e898ef2 (MD5) / Made available in DSpace on 2013-09-11T13:28:23Z (GMT). No. of bitstreams: 1 dissertacao-MiltonSaito.pdf: 8406128 bytes, checksum: 2c7b006b9cca9d097d0ee1116e898ef2 (MD5) Previous issue date: 2013-08-14 / Essa dissertação tem como objetivo a modelagem, implementação e a simulação de um mercado de leilão duplo artificial utilizando a abordagem multiagente. Mercado de leilão duplo permite que ofertas de compra e de venda possam ser feitas a qualquer momento e de forma contínua. As bolsas de Nova Iorque (NYSE) e de Chicago (CME) são exemplos deste tipo de leilão. Mercados artificiais são modelos que têm o objetivo de capturar as propriedades dos mercados reais para reproduzir e analisar a dinâmica do mercado através de experimentos computacionais. Assim como no mercado real, o modelo propõe que os agentes interagem entre si assincronamente em sessões de negociações contínuos. Estas últimas características do modelo são viabilizadas através do uso de técnicas e arcabouços tecnológicos que são atualmente utilizados nos mercados reais. Neste trabalho, são investigados os comportamentos do mercado artificial para diferentes grupos de agentes e parâmetros. Ao longo dos experimentos foram constatados que o volume de negociação e a volatilidade dos preços, por exemplo, são diretamente proporcionais ao orçamento dos agentes. Também foram identificados alguns fatos estilizados nas séries de preços geradas a partir do mercado artificial. / The purpose of this work is to model, implement and simulate a double auction artificial market using a multi-agent approach. Double auction markets allow offers to buy and sell to be made at any time and in a continuous fashion. The New York Stock Exchange (NYSE) and the Chicago Mercantile Exchange (CME) are examples of exchanges that adopt this type of auction. Artificial markets are models that aim to capture the properties of real markets to reproduce and analyze the market dynamics through computational experiments. As in real markets, the model proposes that agents interact asynchronously in continuous trading sessions. These features are made possible by using modern techniques and technological frameworks currently used in real markets. In this work, we investigated the behavior of a artificial market using different groups of agents and parameters. Throughout the experiments it was identified that, for example, trading volume and price volatility are directly proportional to the agents' budget. Also, some stylized facts were identified on the price series generated by the artificial market

因應電子化交易興起而進行的一系列人機互動研究顯示, 縱使人類會透過學習而改善其表現, 電腦化的交易程式獲利能力還是遠勝於真人交易者之表現。本研究遂以遺傳規劃演算法作為學習型交易者之代表, 與一系列電腦化交易策略相競爭, 以探討學習的功效及其限制。 本研究採用離散型雙方喊價機制, 摒除了計算能力所造成之決策時間差異所會帶來的影響, 亦排除掉人類情緒、預期、相關知識不足等可能因子, 在計算能力對等的情況下, 單純地來評估學習與理性設計策略的結果。並且首次嘗試將影響學習至鉅的智商因子帶入模型之中, 實驗結果顯示學習具有相當的能力, 即使是在對環境缺乏認識的情況下, 隨著時間的經過其表現最終可凌駕理性設計的策略之上, 然而學習所需的時間是學習型交易者的一大弱點。同時, 本研究也顯示對於以遺傳規劃建構的學習型交易者而言, 其虛擬智商的參數愈高, 學習的效果也愈佳。此研究因此可作為未來在代理人基經濟學模型中, 更深入地探討智商水準不同所造成之行為差異的基礎。 / The study of a series of human-agent interactions as well as computerized trading tournaments in double auction markets has exhibited a general superiority of computerized trading strategies over learning agents. The ineffectiveness of learning motivates the study of learning versus designed trading agents in this research. We therefore initiates a series of experiments to test the capability of learning GP agents and rationally-designed trading strategies. The results shows that with the cost of time, eventually learning agents can beat all other trading strategies. At the same time, the notion of intelligence is introduced into the model to investigate the influence of individual intelligence on learning ability. We utilize the population size of the GP trader as the proxy variable of IQ which is a measure of general intelligence. The results show that individuals with higher intelligence can perform better than those with lower intelligence, which manifests its importance discovered in Psychological research.

近年來，網際網路(Internet)快速發展，已成為一個無疆無界無時差的市場，如何不被這股潮流所淘汰，我們所提出的解決方案—軟體代理人(software agent)，一位具有人工智慧(artificial intelligence)演化調適(adaptive)能力的代理人，現在已經有許多企業與資訊、管理、電腦科學等各方面專家結合，開始使用軟體代理人來代勞，試想一位永不停止、具有創新、學習適應的員工，企業家可以隨意複製或刪除，隨時配合市場規模，不必擔心任何裁員的負擔，這樣的代理人的問世，勢必對我們的經濟環境帶來莫大的衝擊。 電子商務(electronic commerce)已經行之有年，人類的消費型態似乎不易於因這個轉變而有所改變，消費者如果沒有經過視覺、觸覺、嗅覺等感官的刺激，很難有購買的動機，再加上授信制度的不健全使得電子商務的施行充滿了風險。雖然有這麼多問題，我們仍無法阻擋這股趨勢，在電子科技的進步，3D數位影像、各種感官刺激的傳送、或如同期貨市場上明確公認的規格、法律的修訂、完善的認證制度，接下來我們就是要看軟體代理人的表現。 我們將軟體代理人運用在人工雙方喊價(artificial double auction)的市場，就像真實市場已經有人開始使用自動下單或自動議價代理人的機制一樣。然而市場上是否有必然不敗的策略呢？本文就是要解開這個答案。再進一步來看，待真實市場每個成員受不了生存競爭的壓力，也採取使用代理人的演化性策略，屆時我們的人工市場就是真實市場的縮影，我們在本文也會針對這樣一個具有未來前瞻性市場的表現如何？透過經濟學的角度來揭露市場的本質是否仍然維持？ 在本文軟體代理人即為議價代理人(bargaining agent)，她可以在穩定的(stable)市場環境(其他參與者使用固定策略)中辨別出一些有利的市場特徵，藉由這些特徵發展出有利的策略，而其結果甚至不是很容易想到的策略；接著若每個人都使用議價代理人在市場上交易，這裡我們使用一種納許式過程(Nash-like process)來詮釋，之後再分別依市場的分配效率、價格效率、及所得分配來討論市場績效。