Dissertations / Theses on the topic 'Emissions trading Mathematical models'

Our study examines some design alternatives for a carbon market by exploring the fledgling Alberta carbon market. We attempt to evaluate the performance of these designs on the bases of trade volume, cost efficiency and stability. To achieve this we construct an empirically-calibrated but simple agent-based model, certain aspects of which we selectively modify to incorporate various design options. We make comparisons among these options based on data simulated from the ensuing family of models. We find strong evidence that in general, market design features such as source-of-credits, the scale of the market, and pricing-mechanism are very important considerations that influence the performance of the market. In addition, we find support for the notion that the level of the price cap relative to the average cost of abatement in the market matters, and beyond a threshold, higher price caps are associated with lower levels of performance.
vii, 155 leaves ; 29 cm

Research conducted in mathematical finance focuses on the quantitative modeling of financial markets. It allows one to solve financial problems by using mathematical methods and provides understanding and prediction of the complicated financial behaviors. In this thesis, efforts are devoted to derive and extend stochastic optimization models in financial economics and establish practical algorithms for representing and solving problems in mathematical finance. An option gives the holder the right, but not the obligation, to buy or sell an underlying asset at a specified strike price on or before a specified date. In this thesis, a valuation model for a perpetual convertible bond is developed when the price dynamics of the underlying share are governed by Markovian regime-switching models. By making use of the relationship between the convertible bond and an American option, the valuation of a perpetual convertible bond can be transformed into an optimal stopping problem. A novel approach is also proposed to discuss an optimal inventory level of a retail product from a real option perspective in this thesis. The expected present value of the net profit from selling the product which is the objective function of the optimal inventory problem can be given by the actuarial value of a real option. Hence, option pricing techniques are adopted to solve the optimal inventory problem in this thesis. The goal of risk management is to eliminate or minimize the level of risk associated with a business operation. In the risk measurement literature, there is relatively little amount of work focusing on the risk measurement and management of interest rate instruments. This thesis concerns about building a risk measurement framework based on some modern risk measures, such as Value-at-Risk (VaR) and Expected Shortfall (ES), for describing and quantifying the risk of interest rate sensitive instruments. From the lessons of the recent financial turmoils, it is understood that maximizing profits is not the only objective that needs to be taken into account. The consideration for risk control is of primal importance. Hence, an optimal submission problem of bid and ask quotes in the presence of risk constraints is studied in this thesis. The optimal submission problem of bid and ask quotes is formulated as a stochastic optimal control problem. Portfolio management is a professional management of various securities and assets in order to match investment objectives and balance risk against performance. Different choices of time series models for asset price may lead to different portfolio management strategies. In this thesis, a discrete-time dynamic programming approach which is flexible enough to deal with the optimal asset allocation problem under a general stochastic dynamical system is explored. It’s also interesting to analyze the implications of the heteroscedastic effect described by a continuous-time stochastic volatility model for evaluating risk of a cash management problem. In this thesis, a continuous-time dynamic programming approach is employed to investigate the cash management problem under stochastic volatility model and constant volatility model respectively.
published_or_final_version
Mathematics
Doctoral
Doctor of Philosophy

Two approaches, one using an atmospheric diffusion model and the other an atmospheric tracer, were used to predict the source strength of trace gases from observations of the downwind concentration field. Both approaches do not disturb the prevailing environmental and physical conditions nor the existing biogenic processes. An analytical solution to the advection-diffusion equation was used to back-calculate the source strength from the downwind concentration measurements of (i) single and multipoint (4 and 16 points) trace gas (sulphur hexafluoride (SF6) and methane (CH4)) release experiments conducted over microplots over an open field, and (ii) single point source SF6 release experiments conducted over a forested terrain. Best predictions of the source strength (to within +/-20%) were obtained from concentration observations made along the centreline of the diffusing plumes with the predictions improving when observations at the mean plume height were used. The diffusion model was then used to compute footprint estimates for neutral and unstable conditions, for tower and aircraft based observation platforms above the forest. They showed spatially constrained footprints in the surface layer, due to effective vertical coupling, so that observations from towers and low flying aircraft must be expected to be very site specific, and scaling up to larger areas will have to be done with careful consideration of surface mosaics. Above-canopy sampling of trace gases to determine volatile organic compound emissions were then interpreted in terms of footprint considerations. This was accomplished by defining the upwind canopy areas effectively sampled under the given wind and stability conditions. The analysis demonstrated, for example, that the variability observed in measured isoprene fluxes could be accounted for by varying numbers of randomly distributed clumps of emitter species within a varying footprint. It suggested that heterogeneity of the forest canopy, in ter
Sulphur hexafluoride was also used as an atmospheric tracer in order to estimate CH4 emissions from manure slurry and cattle housed in barns and feedlots. (Abstract shortened by UMI.)

Hedge funds report performance information voluntarily. When they stop reporting they are transferred from the "live" pool of funds to the "defunct" pool. Consequently, liquidated funds constitute a subset of the defunct pool. I present models of hedge fund survival, attrition, and survivorship bias based on liquidation alone. This refines estimates of predictor variables in models of survival, leads to attrition rates of hedge funds to be roughly one half those previously thought, and produces larger estimates of survivorship bias. Survival models based on liquidated funds only, lead to an increase in survival time of 50 to 100 percent relative to survival based on all defunct funds.
In addition to refining estimates of survival time, it is useful to examine how the double fee structure of hedge funds and Commodity Trading Advisors (CTA) affects the incentives of their managers. Young CTAs are usually very small --- they hold few financial assets --- and may not meet their operating expenses with their management fee alone, so their incentive is to take on risk and post good returns. As they grow, their incentive to take on risk diminishes. CTAs in their fifth year diminish their volatility by 25 percent relative to their first year, and diminish returns by 70 percent. We find CTAs to behave more like indexers as they grow, concerned with more with capital preservation than asset management.
Operational risk is a major cause of hedge fund and CTA liquidation. In the banking industry, regulators have called upon institutions to develop models for measuring capital charge for operational losses, and to subject these models to stress testing. Losses are found to be inversely related to GDP growth, and positively related to unemployment. Since losses are thus cyclical, one way to stress test models is to calculate capital charge during good and bad economic regimes. We find loss distributions to have thicker tails during bad regimes. One implication is that banks will likely need to increase their capital charge when economic conditions deteriorate.

We propose a systematic framework for designing adaptive trading strategies that minimize both the mean and the variance of the execution costs. This is achieved by diversifying risk over sequential decisions in discrete time. By incorporating previous trading performance as a state variable, the framework can dynamically adjust the risk-aversion level for future trading. This incorporation also allows the framework to solve the mean-variance problems for different risk aversion factors all at once. After developing this framework, it is then applied to solve three algorithmic trading problems. The first two are trade scheduling problems, which address how to split a large order into sequential small orders in order to best approximate a target price – in our case, either the arrival price, or the Volume-Weighed-Average-Price (VWAP). The third problem is one of optimal execution of the resulting small orders by submitting market and limit orders. Unlike the tradition in both academia and industry of treating the scheduling and order placement problems separately, our approach treats them together and solves them simultaneously. In out-of-sample tests, this unified strategy consistently outperforms strategies that treat the two problems separately.

In dieser Arbeit untersuchen wir die Struktur von Gausschen Prozessen, die durch gewisse lineare Transformationen von zwei Gausschen Martingalen erzeugt werden. Die Klasse dieser Transformationen ist durch nanzmathematische Gleichgewichtsmodelle mit heterogener Information motiviert. In Kapital 2 bestimmen wir für solche Prozesse, die zunächst in einer erweiterten Filtrierung konstruiert werden, die kanonische Zerlegung als Semimartin-gale in ihrer eigenen Filtrierung. Die resultierende Drift wird durch Volterra-Kerne beschrieben. Insbesondere charakterisieren wir diejenigen Prozesse, die in ihrer eigenen Filtrierung eine Brownsche Bewegung bilden. In Kapital 3 konstruieren wir neue orthogonale Zerlegungen der Brownschen Filtrierungen. In den Kapitaln 4 bis 6 wenden wir unsere Resultate zur Charakterisierung Brownscher Bewegungen im Kontext nanzmathematischer Modelle an, in denen es Marktteilnehmer mit zusätzlicher Insider-Information gibt. Wir untersuchen Erweiterungen eines Gleichgewichtsmodells von Kyle [42] und Back [7], in denen die Insider-Information in verschiedener Weise durch Gaussche Martingale spezifiziert wird. Insbesondere klären wir die Struktur von Insider-Strategien, die insofern unaufallig bleiben, als sich die resultierende Gesamtnachfrage wie eine Brownsche Bewegung verhält.
In this thesis, we study Gaussian processes generated by certain linear transformations of two Gaussian martingales. This class of transformations is motivated by nancial equilibrium models with heterogeneous information. In Chapter 2 we derive the canonical decomposition of such processes, which are constructed in an enlarged ltration, as semimartingales in their own ltration. The resulting drift is described in terms of Volterra kernels. In particular we characterize those processes which are Brownian motions in their own ltration. In Chapter 3 we construct new orthogonal decompositions of Brownian ltrations. In Chapters 4 to 6 we are concerned with applications of our characterization results in the context of mathematical models of insider trading. We analyze extensions of the nancial equilibrium model of Kyle [42] and Back [7] where the Gaussian martingale describing the insider information is specified in various ways. In particular we discuss the structure of insider strategies which remain inconspicuous in the sense that the resulting cumulative demand is again a Brownian motion.

[Truncated abstract] The cochlea presumably possesses a number of regulatory mechanisms to maintain cochlear sensitivity in the face of disturbances to its function. Evidence for such mechanisms can be found in the time-course of the recovery of CAP thresholds during experimental manipulations, and in observations of slow oscillations in cochlear micromechanics following exposure to low-frequency tones (the “bounce phenomenon”) and other perturbations. To increase our understanding of these oscillatory processes within the cochlea, and OHCs in particular, investigations into cochlear regulation were carried out using a combination of mathematical modelling of the ionic and mechanical interactions likely to exist within the OHCs, and electrophysiological experiments conducted in guinea pigs. The electrophysiological experiments consisted of electrocochleographic recordings and, in some cases, measurement of otoacoustic emissions, during a variety of experimental perturbations, including the application of force to the cochlear wall, exposure to very-low-frequency tones, injection of direct current into scala tympani, and intracochlear perfusions of artificial perilymph containing altered concentrations of potassium, sodium, and sucrose. To obtain a panoramic view of cochlear regulation under these conditions, software was written to enable the interleaved and near-simultaneous measurement of multiple indicators of cochlear function, including the compound action potential (CAP) threshold, amplitude and waveshape at multiple frequencies, the OHC transfer curves derived from low-frequency cochlear microphonic (CM) waveforms, distortion-product otoacoustic emissions (DPOAEs), the spectrum of the round-window neural noise (SNN), and the endocochlear potential (EP). ... The mathematical model we have developed provided a physiologically-plausible and internally-consistent explanation for the time-courses of the cochlear changes observed during a number of different perturbations. We show that much of the oscillatory behaviour within the cochlea is consistent with underlying oscillations in cytosolic calcium concentration. We conclude that a number of the discrepancies between the simulation results and the experimental data can be resolved if the cytosolic calcium functions as two distinct pools: one which controls basolateral permeability and one which controls slow motility. This two-calcium-pool model is discussed.

We improved the assimilated daily inversion method by conducting model simulation, satellite retrieval, and inverse modeling sequentially on a daily basis. The improved procedure was applied to GOME-2 and OMI NO₂ measurements over China in 2011, respectively. The new daily retrieval-inversion method significantly reduced the systematic bias in inverse modeling of NOₓ emission between using GOME-2 and OMI measurements, and detected more clear seasonal and weekly variations. OMI instrument observed NO₂ columns over China from 2005 to 2010 were analyzed in order to estimate the top-down anthropogenic NOₓ emission trends. The estimated average emission trend is slower than the trend reported for previous years. We find large regional, seasonal, and urban-rural variations in emission trends. These results appear to suggest that a number of factors have significantly reduced or even reversed the increasing trend of NOₓ emissions in more economically developed megacities and southern coastal regions, but their effects are not as significant in other major cities or less economically developed regions. A 1-D chemical transport model was applied to analyze OH and HO₂ radical observations during the Pacific Atmospheric Sulfur Experiment (PASE) near Christmas Island (Kiritimati, 1.52°N 157.24°W) from Aug. 2 through Sep. 10, 2007. In two of fourteen research flights, significantly higher HO₂/OH ratios in the buffer layer than the other flights were found. Model simulations indicated that fast-reacting oxygenated volatile organic compounds, which can react rapidly with OH and provide additional primary radical sources through photolysis, were necessary to explain the observations. During or right before these two flights, the WRF model simulated two strongest shallow convective events during this experiment, suggesting a transport pathway of ocean organics into the buffer layer. Ocean upwelling driven by atmospheric pressure depression during convection may expedite the release of ocean organics.

Recent studies of the environmental Kuznets curve raise questions regarding the relationship between environmental indicators and GDP and the fundamental reasons that explain this relationship. In response, this thesis presents one-sector and two-sector models to analyze the alternative causal relationships between an environmental indicator and GDP at different stages of economic development. These models analyze how economic scale, technology, preferences, and economic structure influence the causality and shape of the relationship. These theoretical studies are followed by two empirical studies. The first tests the causal relationship between CO2 emissions and GDP in Canadian manufacturing industries. The second explores several factors as the fundamental causes that influence the CO2 emissions in the same industries. Factors, such as economic scale, preferences, technological progress, structural change, and energy input, are found to be crucial in the determination of CO2 emissions. The empirical results are positive, but there are data limitations. The empirical studies can be re-evaluated as more data becomes available.

The world bond market is nearly twice as large as the equity market. The goal of this dissertation is to study the dynamics of bond price. Among the liquidity risk, interest rate risk and default risk, this dissertation will focus on the liquidity risk and trading strategy. Under the mathematical frame of stochastic control, we model price setting in U.S. bond markets where dealers have multiple instruments to smooth inventory imbalances. The difficulty in obtaining the optimal trading strategy is that the optimal strategy and value function depend on each other, and the corresponding HJB equation is nonlinear. To solve this problem, we derived an approximate optimal explicit trading strategy. The result shows that this trading strategy is better than the benchmark central symmetric trading strategy.
ID: 029809224; System requirements: World Wide Web browser and PDF reader.; Mode of access: World Wide Web.; Thesis (Ph.D.)--University of Central Florida, 2011.; Includes bibliographical references (p. 101-103).
Ph.D.
Doctorate
Mathematics
Sciences

La prima parte della Tesi si occupa dello studio delle strategie di investimento ottime nel settore energetico e in ricerca e sviluppo, nell'ambito di politiche di stabilizzazione dei gas serra nell'atmosfera. La seconda parte tratta invece metodi per la quantificazione degli impatti economici dei cambiamenti climatici sul settore agricolo, considerando tutte le possibilità di adattamento.
The first part of the Thesis discusses optimal investment strategies in the energy sector and in R&D for knowledge advancements to stabilize atmospheric concentrations of GHG. The second part deals instead with the measurement of impacts of climate change on agriculture considering all possible adaptation options.

This thesis presents a mathematical framework to model trading of financial assets on an exchange. The interaction between agents on the exchange is modeled as the Nash equilibrium of a demand schedule auction. The submission of demand schedules in the auction is meant to proxy for the submission of limit and market orders on an exchange. Chapter 1 considers this auction in a one-period setting, highlighting the importance of noisy flow for obtaining a unique Nash equilibrium.

Chapter 2 is the core of the thesis and considers the auction in a continuous time setting. Here the agents trading on the exchange have quadratic-type preferences, and in equilibrium they must clear an exogenously specified stream of market orders. Chapter 3 considers alternative and more realistic dynamics for the exogenous market orders. Chapter 4 endogenizes the market orders by considering an agent executing orders on behalf of noisy clients.. Chapter 5 considers the same model as in Chapter 2, except with a consumption based utility function for each agent.

Yung Yan Keung.
Thesis (M.Phil.)--Chinese University of Hong Kong, 1999.
Includes bibliographical references (leaves 144-147).
Chapter Chapter 1: --- Introduction --- p.1
Chapter 1.1 --- Review on Portfolio Theory --- p.3
Chapter - 1.1.1 --- Expected Return and Risk of a Security --- p.3
Chapter -1.1.2 --- Expected Return and Risk of a Portfolio --- p.4
Chapter -1.1.3 --- The Feasible Set --- p.5
Chapter - 1.1.4 --- Assumptions on the Investor --- p.6
Chapter -1.1.5 --- Efficient Portfolios --- p.6
Chapter -1.1.5.1 --- Bounds on the Return and Risk of a portfolio --- p.6
Chapter -1.1.5.2 --- Concavity of the Efficient Set --- p.8
Chapter -1.1.6 --- The Market Model --- p.9
Chapter -1.1.7 --- Risk-free Asset --- p.11
Chapter - 1.1.8 --- Portfolio involving Risk-free Asset --- p.12
Chapter -1.1.9 --- The Sharpe Ratio --- p.14
Chapter 1.2 --- Review on Some Trading Models --- p.19
Chapter -1.2.1 --- Buy and Hold Model --- p.19
Chapter -1.2.2 --- Trading Model with Prediction Criteria --- p.20
Chapter -1.2.2.1 --- Two School of Theories --- p.20
Chapter - 1.2.2.2 --- Prediction of the stock price movement --- p.20
Chapter -1.2.2.3 --- The Use of Neural Network in Prediction --- p.21
Chapter -1.2.2.4 --- Single Step and Multi-step Prediction --- p.23
Chapter - 1.2.2.5 --- Trading Model based on Prediction Criteria --- p.25
Chapter - 1.2.2.6 --- For More Accurate Prediction --- p.25
Chapter -1.2.3 --- Weigend's Model --- p.26
Chapter - 1.2.3.1 --- Introduction --- p.26
Chapter -1.2.3.2 --- The Model Setup --- p.26
Chapter -1.2.3.3 --- The Objective Functions --- p.27
Chapter - 1.2.3.4 --- The Gradient Ascending Algorithm --- p.27
Chapter -1.2.3.5 --- The Gradient of the Sharpe Ratio --- p.27
Chapter - 1.2.3.6 --- The Training Procedure --- p.28
Chapter - 1.2.3.7 --- Some Properties of the Sharpe Ratio Training --- p.28
Chapter -1.2.4 --- Bengio's Model --- p.29
Chapter -1.2.4.1. --- Overview --- p.29
Chapter -1.2.4.2. --- The Trading System --- p.29
Chapter - 1.2.4.3 --- The Objective Function: the Portfolio Return --- p.31
Chapter - 1.2.4.4. --- The Training Process --- p.32
Chapter - 1.2.4.5 --- Computer Simulation --- p.34
Chapter - 1.2.4.6 --- Discussion --- p.36
Chapter Chapter 2: --- The Naive Sharpe Ratio Model --- p.38
Chapter - 2.1 --- Introduction --- p.39
Chapter - 2.2 --- Definition of the Naive Sharpe Ratio --- p.39
Chapter - 2.3 --- Gradient of Naive Sharpe Ratio with respect to the portfolio weighting: --- p.40
Chapter - 2.4 --- The Training Process --- p.40
Chapter - 2.5 --- Analysis of the Gradient --- p.41
Chapter -2.6 --- Compare with Bengio's and Weigend's Model --- p.42
Chapter -2.7. --- Computer Simulations --- p.43
Chapter -2.7.1 --- Experiment 1: How the Sharpe Ratio is Maximized --- p.43
Chapter -2.7.1.1 --- Experiment 11 --- p.44
Chapter -2.7.1.2 --- Experiment 12 --- p.45
Chapter -2.7.1.3 --- Experiment 13 --- p.46
Chapter -2.7.2 --- Experiment 2: Reducing the Unique Risk --- p.49
Chapter -2.7.3 --- Experiment 3: Apply to the Stock Market --- p.52
Chapter -2.8 --- Redefining the Naive Sharpe ratio with down-side risk --- p.56
Chapter -2.8.1 --- Definitions --- p.56
Chapter -2.8.2 --- Gradient of the Downside Nai've Sharpe Ratio --- p.57
Chapter -2.8.3 --- Analysis of the gradient of the new Sharpe ratio --- p.57
Chapter -2.8.4 --- Experiment: Compared with Symmetric Risk --- p.59
Chapter -2.8.4.1 --- Experimental Setup --- p.59
Chapter -2.8.4.2 --- Experimental Result --- p.60
Chapter -2.8.4.3 --- Discussion --- p.62
Chapter - 2.9 --- Further Discussion --- p.63
Chapter Chapter 3: --- The Total Sharpe Ratio Model --- p.64
Chapter - 3.1 --- Introduction --- p.65
Chapter -3.2 --- Defining risk of portfolio in terms of component securities' risk --- p.65
Chapter -3.2.1. --- Return for Each Security and the Whole Portfolio at Each Time Step --- p.65
Chapter -3.3.2. --- Covariance of the Individual Securities' Returns --- p.66
Chapter -3.2.3. --- Define the Sharpe Ratio and the Objective Function --- p.66
Chapter -3.2.3.1. --- The Excess Return --- p.66
Chapter -3.2.3.2. --- The Risk --- p.67
Chapter -3.2.3.3. --- The Sharpe Ratio at Time t --- p.67
Chapter -3.2.3.4. --- The Objective Function: the total Sharpe ratio --- p.67
Chapter -3.2.3.5. --- The Training Process --- p.68
Chapter -3.3 --- Calculating the Gradient of the Total Sharpe Ratio --- p.69
Chapter -3.4. --- Analysis of the Total Sharpe Ratio Gradient --- p.70
Chapter -3.4.1 --- The Gradient Vector of the Sharpe Ratio at a Particular Time Step --- p.70
Chapter -3.4.2 --- The Gradient Vector of the Risk --- p.70
Chapter - 3.5 --- Computer Simulation: --- p.72
Chapter -3.5.1 --- Apply to the Stock Market1 --- p.72
Chapter -3.5.1.1 --- Objective --- p.72
Chapter - 3.5.1.2 --- Experimental Setup --- p.72
Chapter -3.5.1.3 --- The Experimental Result --- p.73
Chapter -3.5.2 --- Apply to the Stock Market2 --- p.78
Chapter -3.5.2.1 --- Objective --- p.78
Chapter -3.5.2.2 --- Experimental Setup --- p.78
Chapter -3.5.2.3 --- The Experimental Result --- p.79
Chapter -3.6 --- Defining the Total Sharpe Ratio in terms of Downside Risk --- p.84
Chapter - 3.6.1. --- Introduction --- p.84
Chapter -3.6.2. --- Covariance of the individual securities' returns --- p.84
Chapter -3.6.3. --- Define the Downside Risk Sharpe ratio and the objective function --- p.85
Chapter -3.6.3.1. --- The Excess Return --- p.85
Chapter -3.6.3.2. --- The Downside Risk --- p.85
Chapter -3.6.3.3. --- The Sharpe ratio at time T --- p.85
Chapter -3.6.3.4. --- The Objective function --- p.85
Chapter -3.6.4. --- The Training Process --- p.85
Chapter -3.7 --- Total Sharpe Ratio involving Transaction Cost --- p.86
Chapter -3.7.1 --- Introduction --- p.86
Chapter -3.7.2 --- Return for each stock and the whole portfolio at each time step --- p.86
Chapter -3.7.3 --- Linear Approximation of the Portfolio's return --- p.88
Chapter -3.7.4 --- Covariance of the individual securities' returns --- p.89
Chapter -3.7.5 --- Define the Sharpe ratio and the objective function --- p.90
Chapter -3.7.5.1 --- The Excess Return --- p.90
Chapter -3.7.5.2 --- The Risk --- p.90
Chapter -3.7.5.3 --- The Sharpe Ratio at time T --- p.90
Chapter -3.7.5.4 --- The Objective Function --- p.90
Chapter -3.7.6 --- Calculation of the gradient of the Total Sharpe ratio --- p.91
Chapter -3.7.7. --- Analysis of the Total Sharpe Ratio Gradient --- p.94
Chapter -3.7.7.1 --- The Gradient Vector of the Sharpe Ratio at a Particular Time Step --- p.94
Chapter -3.7.7.2 --- The Gradient Vector of the Risk --- p.94
Chapter -3.7.8 --- Experiment 1: Compare with Buy and Hold Method --- p.96
Chapter -3.7.8.1 --- Experiment 11 --- p.96
Chapter -3.7.8.2. --- Experiment 12 --- p.102
Chapter -3.7.9 --- Experiment 2: Compared with Naive Sharpe Ratio --- p.108
Chapter -3.7.9.1 --- Objective --- p.108
Chapter -3.7.9.2. --- Experimental Setup --- p.108
Chapter -3.7.9.3. --- The Experimental Result --- p.109
Chapter - 3.7.10 --- Experiment 3: Compared with other models --- p.113
Chapter - 3.7.10.1 --- Experiment 31 --- p.113
Chapter - 3.7.10.2. --- Experiment 32 --- p.117
Chapter -3.7.11 --- Experiment 4: Apply to the Stock Market --- p.121
Chapter -3.7.11.1 --- Objective --- p.121
Chapter - 3.7.11.2. --- Experimental Setup --- p.121
Chapter -3.7.11.3. --- The Experimental Result --- p.121
Chapter Chapter 4: --- Conclusion --- p.126
Appendix A --- p.130
Appendix B --- p.139
Appendix C --- p.141
Appendix D --- p.142
Reference --- p.144

在現今的社會，協整技術已被廣泛應用於金融和計量經濟領域，特別用於構建股票市場的統計套利策略。在這一篇論文中，我們主要考察在衍生品市場中，基於協整技術的套利交易策略，這一策略的主要研究對象是隱含波動率。利用隱性波動率的線性組合的均值回歸的特性，通過配對兩隻帶有正利差(如theta) 的短期平價歐式跨式期權來獲利。同時，構建實際波動率的模型和預測未來實際波動率的模型將會用於補充這一交易策略的不足，隱性一實際條件和Gamma-Vega條件被引入來提高交易策略的效率。這一策略的績效分析是基於三年的歷史外匯期權數據。從實證數據中，基於協整技術的策略能賺取利潤，而且Vega在利潤中起著重要的作用，並且無論是隱性一實際條件還是Gamma-Vega條件都是有效的。
The notion of cointegration has been widely used in finance and econometrics, in particular in constructing statistical arbitrage strategies in the stock market. In this thesis, an arbitrage trading strategy for derivatives based on cointegration is studied to account for the volatility factor. Pairs of short dated at-the-money straddles of European options with positive net carry (i.e. theta) are used to capture the mean-reverting property of the linear combinations of implied volatilities. Furthermore, modeling and forecasting realized volatility are also considered as a supplement to the trading strategy. Implied-Realized Criertion and Gamma-Vega Criterion are introduced to improve the trading strategy. A performance analysis is conducted with a 3-year historical data of Foreign Exchange Options. From the empirical results, the portfolio based on the cointegration strategy makes a profit, where Vega plays a dominant role, and either the Implied-Realized Criertion or the Gamma-Vega Criterion is effective.
Detailed summary in vernacular field only.
Pun, Lai Fan.
Thesis (M.Phil.)--Chinese University of Hong Kong, 2013.
Includes bibliographical references (leaves 43-45).
Abstracts also in Chinese.
List of Tables --- p.v
List of Figures --- p.vi
Chapter 1 --- Introduction --- p.1
Chapter 2 --- Basic Ideas --- p.4
Chapter 2.1 --- Cointegration and Johansen’s Methodology --- p.4
Chapter 2.1.1 --- Cointegration --- p.4
Chapter 2.1.2 --- Johansen’s Methodology --- p.5
Chapter 2.2 --- Cointegration Pairs Trading Strategy --- p.6
Chapter 2.3 --- Modelling and Forecasting Realized Volatility --- p.8
Chapter 3 --- Cointegration Pairs Trading Strategy On Derivatives --- p.10
Chapter 3.1 --- Trading On Implied Volatility --- p.10
Chapter 3.2 --- Cointegration Trading Strategy --- p.12
Chapter 3.3 --- Greek Letters --- p.13
Chapter 3.3.1 --- Requirements of the Trade --- p.13
Chapter 3.3.2 --- Approximation of the Expected P/L --- p.15
Chapter 3.4 --- Foreign Exchange Options --- p.18
Chapter 3.4.1 --- Cointegration Pairs --- p.19
Chapter 3.4.2 --- Trading Process --- p.21
Chapter 3.4.3 --- More Examples --- p.22
Chapter 4 --- Further Trading Strategies --- p.26
Chapter 4.1 --- Estimation of Realized Volatility --- p.26
Chapter 4.2 --- Implied-Realized Criterion --- p.27
Chapter 4.3 --- Gamma-Vega Criterion --- p.29
Chapter 4.4 --- Summary --- p.32
Chapter 5 --- Conclusion and Further Discussion --- p.37
A --- p.39
B --- p.41
Bibliography --- p.43

In order to meet the US EPA's National Ambient Air Quality Standards as set under the provisions of the Clean Air Act, states and regions throughout the United States are designing cap and trade programs aimed at reducing the emissions of the two dominant precursors for ozone, nitrogen oxides (NOx) and Volatile Organic Compounds (VOCs). While emission cap and trade programs are becoming more common, relatively few analyses have examined the air quality implications of moving emissions from one location to another (due to trading of emissions between facilities), from one sector to another (due to the use of technologies such as Plug-in Electric Hybrid Vehicles - PHEVs), and changing the temporal distribution of emissions (through emissions trading among facilities with different temporal profiles). This thesis will examine, in detail, the air quality implications of two emission cap and trade programs. The first program is a NOx trading program that covers Electricity Generating Units (EGUs) in the Northeastern United States. Results show that refining the temporal limits on this cap and trade program, by charging facilities more to emit NOx on days when ozone is most likely to form, has the potential to significantly reduce NOx emissions and ozone concentrations. Additionally, this research also shows that, for this region, the spatial redistribution of NOx emissions due to trading leads to greater ozone reductions than similar amounts of NOx emission reductions applied evenly across all facilities. Analyses also indicate that displacing emissions from the on-road mobile sector (vehicles) to the EGU sector through the use of PHEVs decreases ozone in most areas, but some highly localized areas show increases in ozone concentration. The second trading program examined in this thesis is limited to Houston, Texas, where a VOC trading program is focused on a sub-set of four Highly Reactive Volatile Organic Compounds (HRVOCs), which have been identified as having substantial ozone formation potential. Work presented in this thesis examined whether this trading program, in its current form or in an expanded form, could lead to air pollution hot spots, due to spatial reallocation of emissions. Results show that the program as currently designed is unlikely to lead to ozone hot spots, so no further spatial limitations are required for this program. Expanding the trading to include Other VOCs, fugitive emissions and chlorine emissions, based on reactivity weighted trading, is also unlikely to lead to the formation of ozone hot spots, and could create more flexibility in a trading market that is currently not very active. Based on these air quality modeling results, policy suggestions are provided that may increase participation in the trading market. These case studies demonstrate that use of detailed air analyses can provide improved designs for increasingly popular emission cap and trade programs, with improved understanding of the impacts of modifying spatial and temporal distributions of emissions.
text

Li, Xu.
Thesis submitted in: December 2008.
Thesis (M.Phil.)--Chinese University of Hong Kong, 2009.
Includes bibliographical references (leaves 41-43).
Abstracts in English and Chinese.
Abstract --- p.i
Acknowledgement --- p.ii
Chapter 1 --- Introduction --- p.1
Chapter 2 --- Background Study --- p.6
Chapter 3 --- Model Formulation --- p.8
Chapter 3.1 --- Trading Cost Function --- p.9
Chapter 3.2 --- Notations and Optimality Equation --- p.11
Chapter 4 --- Optimal Policy --- p.14
Chapter 4.1 --- Preliminary Assumption and Results --- p.14
Chapter 4.1.1 --- "Generalized (s, 5, H) Policy" --- p.14
Chapter 4.1.2 --- Polya Distribution and Quasi-K-convex --- p.15
Chapter 4.1.3 --- Assumptions --- p.20
Chapter 4.2 --- Single Period Problem --- p.23
Chapter 4.3 --- Finite-Period Problem --- p.30
Chapter 4.4 --- The Algorithm --- p.36
Chapter 5 --- Conclusion --- p.39
Bibliography --- p.41

Ip Tak Sang.
Thesis (M.Phil.)--Chinese University of Hong Kong, 2006.
Includes bibliographical references (leaves 65-67).
Abstracts in English and Chinese.
Chapter Chapter 1 --- Introduction --- p.1
Chapter Chapter 2 --- Data and Methodology --- p.4
Chapter Chapter 3 --- Results
Chapter 3.1 --- Performance of Long/Short Strategies --- p.11
Chapter 3.2 --- Subsample and Sensitivity Analysis --- p.17
Chapter 3.3 --- Autocorrelation Analysis --- p.25
Chapter Chapter 4 --- Discussion and Conclusion --- p.27
Appendices
Chapter A.1 --- Exchange Rates Figures --- p.28
Chapter A.2 --- Tables --- p.32
References --- p.65

by Fu Yiu-Hang.
Thesis (M.B.A.)--Chinese University of Hong Kong, 1999.
Includes bibliographical references (leaves 66-67).
ABSTRACT --- p.ii
TABLE OF CONTENTS --- p.ii
LIST OF TABLES --- p.vi
LIST OF EXHIBITS --- p.vii
PREFACE --- p.viii
ACKNOWLEDGMENTS --- p.x
Chapter
Chapter I. --- INTRODUCTION --- p.1
What is an Option? --- p.1
Options Market --- p.2
Uses of Options --- p.2
Value of Options --- p.3
Index Options --- p.4
Hang Seng Index Options --- p.4
Chapter II. --- BASIC PROPERTIES OF OPTIONS --- p.5
Assumptions --- p.5
Notation --- p.5
Option Prices at Expiration --- p.6
Call Option Prices at Expiration --- p.6
Put Option Prices at Expiration --- p.6
Upper Bounds for Option Prices --- p.6
Upper Bounds for Call Option Prices --- p.6
Upper Bounds for Put Option Prices --- p.6
Lower Bounds for European Option Prices --- p.7
Lower Bounds for European Call Option Prices --- p.7
Lower Bounds for European Put Option Prices --- p.8
Put-Call Parity --- p.8
Chapter III. --- FACTORS AFFECTING OPTION PRICES --- p.10
Price of Underlying Instrument --- p.10
Exercise Price of the Option --- p.10
Volatility of the Price of Underlying Instrument --- p.11
Time to Expiration --- p.11
Risk-free Rate --- p.11
Dividends --- p.12
Chapter IV. --- OPTION PRICING MODEL --- p.13
Assumptions --- p.13
The Price of Underlying Instrument Follows a Lognormal Distribution --- p.13
The Variance of the Rate of Return of Underlying Instrument is a Constant --- p.17
The Risk-free Rate is a Constant --- p.19
No Dividends are Paid --- p.20
There are No Transaction Costs and Taxes --- p.20
The Black-Scholes Option Pricing Model --- p.21
Notation --- p.21
The Formulas --- p.21
The Variables --- p.22
Properties of the Black-Scholes Formulas --- p.22
Implied Volatility --- p.23
Bias of the Black-Scholes Option Pricing Model --- p.26
Other Option Pricing Models。……………… --- p.27
Chapter V. --- SENSITIVITIES OF OPTION PRICE TO ITS FACTORS --- p.29
Delta --- p.29
Vega --- p.30
Theta --- p.31
Rho --- p.32
Gamma --- p.33
Managing the Change in the Value of Option --- p.34
Sensitivities of Portfolio Value to the Factors --- p.34
Chapter VI. --- TRADING STRATEGIES OF OPTIONS --- p.35
Methodology --- p.35
Limitations --- p.36
Basic Strategies --- p.37
Long Call --- p.37
Short Call --- p.39
Long Put --- p.40
Short Put --- p.42
Spread Strategies --- p.43
Money Spread --- p.43
Ratio Spread --- p.46
Box Spread --- p.46
Butterfly Spread --- p.46
Condor --- p.49
Calendar Spread --- p.49
Diagonal Spread --- p.52
Combination Strategies --- p.52
Straddle --- p.52
Strap --- p.54
Strip --- p.54
Strangle --- p.54
Selecting Trading Strategies Intelligently --- p.56
Chapter VII. --- CONCLUSIONS --- p.57
APPENDICES --- p.60
BIBLIOGRAPHY --- p.66

This thesis examines how public announcement events with different characteristics affect the probability of informed trading (PI). Using the Bollen, Smith and Whaley (2004) model of inferring PI directly from trades, we investigate the differences in PI between the pre-announcement period and post-announcement period from 2002 to 2008 in the American stock market along two dimensions: whether announcements are scheduled; and other characteristics related to the content such as payment methods, offer premium (takeover announcements) and earnings surprises (earnings announcements). We first focus on unscheduled takeover announcements. Our results show that PI (and bid/ask spread) is significantly higher in the pre-announcement period compared to that in the post-announcement period. Further, we link the changes in PI to takeover announcement characteristics. We show that PI is significantly higher in the pre-event period for successful offers and cash offers as well as offers with relatively high premiums. In contrast, the changes in PI after announcement are not significant for unsuccessful offers, stock or mixed offers as well as offers with relatively low premiums. We then investigate informed trading around scheduled earnings announcements. Results indicate that PI (and bid/ask spread) before earnings announcements is not significantly higher than after the announcement. However, when breaking down the sample according to the size of the earnings surprise, we find significant incremental PI in the pre-earnings period when reported earnings contain big surprises with respect to the forecasts. Finally, we contrast PI around public announcements, conditioning on whether the announcement is scheduled or unscheduled. We find that the level of PI for scheduled earnings announcements are significantly higher than for unscheduled takeover announcements in the post-announcement period but not in the pre-announcement period. Our results are consistent with the argument that an announcement that is anticipated stimulates relatively more private information gathering, and hence the degree of information asymmetry around scheduled announcements might be higher. We also find that trading volume dries up in the pre-announcement period for scheduled earnings announcements. -- provided by Candidate.

Leong U Man.
Thesis (M.Phil.)--Chinese University of Hong Kong, 2006.
Includes bibliographical references (leaves 32-34).
Abstracts in English and Chinese.
Chapter 1 --- Introduction --- p.1
Chapter 2 --- Structural Bond Pricing Models --- p.5
Chapter 2.1 --- The Merton Model --- p.5
Chapter 2.2 --- Extended Merton Model --- p.6
Chapter 2.3 --- Longstaff and Schwartz Model --- p.8
Chapter 3 --- Methodology --- p.11
Chapter 3.1 --- Maximum Likelihood Estimation --- p.13
Chapter 3.2 --- Non-linear Filtering Process --- p.13
Chapter 3.3 --- Modification for LS Model --- p.15
Chapter 4 --- Simulation and Empirical Analysis --- p.16
Chapter 4.1 --- Simulation Study --- p.16
Chapter 4.2 --- Empirical Analysis --- p.19
Chapter 4.2.1 --- Bond Selection --- p.19
Chapter 4.2.2 --- Result for EM Model --- p.21
Chapter 4.2.3 --- Result for LS Model --- p.24
Chapter 4.3 --- Implications from Empirical Analysis --- p.28
Chapter 5 --- Conclusion --- p.30
Bibliography --- p.32

Thesis (Ph. D.)--University of Tennessee, Knoxville, 2002.
Title from title page screen (viewed Sept. 25, 2002). Thesis advisor: Gregory D. Reed. Document formatted into pages (xx, 325 p. : ill. (some col.), map). Vita. Includes bibliographical references (p. 173-178).

技術分析包括圖表分析和技術指標分析。比較兩者，前者偏於主觀，並且解讀方式不一，而後者卻能用科學方法來考量。本研究論文先分析市場上流行的技術指標，移動平均線。交易員觀測兩條不同日數的移動平均線，從兩線相交處尋找進出市場的時機。從領域來看，兩條不同日數的移動平均線之差屬於一種帶通濾波器。本文將解釋帶通濾波器與市場進出規則之間的關係。除了移動平均線這種線性方法，我們同時考慮非線性的訊號處理工具。特別地，本研究採用近代提出的經驗模態分解法，得出類似移動平均線相交法的一種新交易策略。我們將文中提及的方法應用在香港及中國過去五年的股票市場，並給出數值結果以顯其效。
Technical analysis includes chart pattern reading and stock market indicators. While the former is subjective and open to different interpretations, the latter is quantied in a more scientic way. The moving average, a popular market indicator, will be analyzed in this thesis. Traders monitor the crossovers of two moving averages with different durations to nd market entry timings. From the viewpoint of frequency domain, the difference of two such moving averages is found to be a band-pass filter. The relation between band-pass filter and market entry strategy is explained. Apartfrom linear methods such as the moving average,non linear signal processing tool is also studied. In particular,the modern empirical mode decomposition is applied to derive a new trading strategy similar to the moving average crossover rule. The introduced methods are put to the test in the Hong Kong and Chinese stock markets for the last five years. Numerical results are presented to show the performance of the methods.
Detailed summary in vernacular field only.
Lee, Tsz Ho.
Thesis (M.Phil.)--Chinese University of Hong Kong, 2012.
Includes bibliographical references (leaves 64-66).
Abstracts also in Chinese.
Chapter 1 --- Introduction --- p.7
Chapter 2 --- Linear Filters --- p.11
Chapter 2.1 --- Introduction --- p.11
Chapter 2.2 --- Frequency response --- p.13
Chapter 2.3 --- Recursive filters --- p.16
Chapter 2.4 --- Convolution theorem --- p.20
Chapter 3 --- Momentum Indicators --- p.23
Chapter 3.1 --- Introduction --- p.23
Chapter 3.2 --- Momentum indicators --- p.24
Chapter 3.3 --- Crossover of two moving averages --- p.25
Chapter 3.4 --- MACD and acceleration indicators --- p.27
Chapter 4 --- Profitability of Momentum Indicators --- p.33
Chapter 4.1 --- Introduction --- p.33
Chapter 4.2 --- Trading methodology --- p.34
Chapter 4.3 --- Evaluating the performance --- p.36
Chapter 4.4 --- Results of evaluation --- p.39
Chapter 5 --- Empirical Mode Decomposition --- p.45
Chapter 5.1 --- Introduction --- p.45
Chapter 5.2 --- Instantaneous frequency --- p.46
Chapter 5.3 --- Empirical mode decomposition --- p.47
Chapter 5.4 --- Trading methodology --- p.50
Chapter 5.5 --- Results of evaluation --- p.52
Chapter 6 --- Discussions --- p.57
Chapter A Descriptive Statistics and Additional Numerical Results --- p.60
Bibliography --- p.64

by Sze Kin Wan.
Thesis (M.Phil.)--Chinese University of Hong Kong, 1998.
Includes bibliographical references (leaves 69-75).
Abstract also in Chinese.
ACKNOWLEDGMENTS --- p.iii
LIST OF TABLES --- p.iv
LIST OF ILLUSTRATIONS --- p.v
CHAPTER
Chapter ONE --- INTRODUCTION --- p.1
Chapter TWO --- REVIEW OF THE LITERATURE --- p.7
Stock Returns and Trading Volume
Volatility
Chapter THREE --- ECONOMETRIC ANALYSIS --- p.16
Unit Root Tests
Lag Length Tests
Causality Detection between Two Series
ARCH Modelling
Chapter FOUR --- DATA AND ESTIMATION RESULTS --- p.34
Data
Unit Root Test
Optimal Lag Length
Causality Detection
GARCH Modelling
Chapter FIVE --- CONCLUSION --- p.62
APPENDIX --- p.67
BIBLIOGRAPHY --- p.69
ILLUSTRATIONS --- p.76

by To Kwok-Fai.
Thesis (M.B.A.)--Chinese University of Hong Kong, 1992.
Includes bibliographical references (leaves 83-84).
ABSTRACT --- p.ii
ACKNOWLEDGEMENTS --- p.iii
TABLE OF CONTENTS --- p.iv
LIST OF ILLUSTRATIONS --- p.vi
LIST OF TABLES --- p.viii
Chapter
Chapter I. --- INTRODUCTION --- p.1
Chapter II. --- THE GROWTH AND CHANGING CHARACTER OF THE FOREIGN EXCHANGE MARKET --- p.3
Three Economic Blocs --- p.3
Increase of Trading Volume --- p.4
Shift In Customer Base --- p.5
Twenty-four Hours Global Market --- p.5
Growth in the Use of Computer --- p.6
Chapter III. --- FORECASTING OF FOREIGN EXCHANGE RATE --- p.7
Efficient Market Hypothesis and Random Walk Theory --- p.7
The Hypothesis --- p.7
Implications --- p.9
Chaos Theory --- p.9
Definition --- p.9
Phenomena in Foreign Exchange Market --- p.9
Implications --- p.12
Fundamental Analysis in Forecasting Foreign Exchange Rate --- p.12
Technical Analysis in Forecasting Foreign Exchange Rate --- p.15
Other Factors Influencing Foreign Exchange Rate --- p.17
Chapter IV. --- METHODOLOGY --- p.18
Collection of Data --- p.18
Selection of Trading Systems --- p.20
Construction of Trading Systems --- p.21
Simple Moving Average Trading System --- p.21
Directional Movement Index Trading System --- p.22
Evaluation of Trading Performance --- p.27
Chapter V. --- RESULTS AND FINDINGS --- p.30
Simple Moving Average Trading System --- p.30
Directional Movement Index Trading System --- p.40
Comparison of the Two Trading Systems --- p.50
Current Net Profit or Loss --- p.50
Sample Standard Deviation --- p.52
Sharpe Ratio --- p.52
Ratio of Average Profit per Profitable Transaction to Average Loss per Losing Transaction --- p.55
Chapter VI. --- CONCLUSIONS --- p.57
APPENDIX
Chapter 1. --- Program Listing of Simple Moving Average Trading System Performance Report --- p.59
Chapter 2. --- Program Listing of Directional Movement Index Trading System Performance Report --- p.63
Chapter 3. --- "Detailed Listing of USD/DEM High, Low and Close Exchange Rate from Oct 18 1988 to Dec 31 1991" --- p.67
BIBLIOGRAPHY --- p.83

This thesis generalizes stochastic portfolio theory in two different aspects. The first part demonstrates the functional generation of portfolios in a pathwise way. This notion of functional generation of portfolios was first introduced by E.R. Fernholz, to construct a variety of portfolios solely in the terms of the individual companies' market weights. I. Karatzas and J. Ruf developed recently another approach to the functional construction of portfolios, which leads to very simple conditions for strong relative arbitrage with respect to the market. Both of these notions of functional portfolio generation are generalized in a pathwise, probability-free setting; portfolio generating functions, possibly less smooth than twice-differentiable, involve the current market weights, as well as additional bounded-variation functionals of past and present market weights. This generalization leads to a wider class of functionally-generated portfolios than was heretofore possible to analyze, and to improved conditions for outperforming the market portfolio over suitable time-horizons. The second part develops portfolio theory in open markets. An open market is a subset of the entire equity market, composed of a certain fixed number of top-capitalization stocks. Though the number of stocks in open market is fixed, the constituents of the market change over time as each company's rank by its market capitalization fluctuates. When one is allowed to invest also in money market, an open market resembles the entire 'closed' equity market in the sense that most of the results that are valid for the entire market, continue to hold when investment is restricted to the open market. One of these results is the equivalence of market viability (lack of arbitrage) and the existence of num\'eraire portfolio (portfolio which cannot be outperformed). When access to the money market is prohibited, the class of portfolios shrinks significantly in open markets. In such a case, we discuss the Capital Asset Pricing Model, how to construct functionally-generated portfolios, and the concept of universal portfolio in open market setting.

Lau, Chi Keung.
Thesis (M.Phil.)--Chinese University of Hong Kong, 2000.
Includes bibliographical references (leaves 187-188).
Abstracts in English and Chinese.
Abstract --- p.i
Acknowledgements --- p.iii
Table of Contents --- p.iv
List of Chosen Samples Results --- p.v
List of Tables --- p.vi
List of Figures --- p.vii
Chapter Chapter 1. --- Introduction --- p.1
Chapter Chapter 2. --- Literature Review --- p.4
Chapter 2.1 --- Real Estate Literature --- p.4
Chapter 2.2 --- Financial Literature --- p.8
Chapter Chapter 3. --- Methodology --- p.15
Chapter 3.1 --- Augmented Dickey Fuller Test --- p.15
Chapter 3.2 --- Band-Pass Filter --- p.18
Chapter Chapter 4. --- Data Description --- p.20
Chapter Chapter 5. --- Empirical Results --- p.23
Chapter 5.1 --- Contemporaneous Correlation --- p.24
Chapter 5.2 --- Results after Band-Pass Filtering --- p.26
Chapter 5.3 --- Lead-lag Relationship Analysis --- p.30
Chapter Chapter 6. --- Conclusion --- p.35
Appendix 1. Variable Definition --- p.38
Appendix 2. Limitation --- p.41
Appendix 3. Results of Chosen Samples --- p.45
Appendix 4. Tables --- p.54
Appendix 5. Figures --- p.109
Bibliography --- p.187

The United States Environmental Protection Agency has estimated that over 111 million people reside in areas that exceed the National Ambient Air Quality Standards for ozone. One major source of the chemical precursors (nitrogen dioxides and volatile organic compounds (VOCs)) for ozone are motor vehicles. The overall goal of this research is to improve the knowledge base related to VOC refueling and evaporative emissions from motor vehicles. Refueling, running loss, hot soak, and diurnal loss total and speciated VOC emissions were investigated. A total of 12 uncontrolled refueling events were completed and involved the determination of volumetric flow rates of gasoline vapor during refueling, as well as total and speciated VOC concentrations. Total VOC emissions were compared with two commonly used algorithms. Speciated VOC vapor profiles were compared with two published gasoline vapor profiles and theoretical predictions based on knowledge of liquid composition and environmental conditions. An evaluation of refueling emissions impacts on ozone formation potentials using MIR was completed and results were compared against speciated emissions and MOBILE-based total VOC emissions estimates coupled with a default speciation profile. Refueling VOC emissions and resultant ozone formation potential may be underestimated in existing emission inventories, particularly during the summer ozone season, A model was developed to predict the speciation of VOCs associated with evaporative emissions from motor vehicles. Model-predicted speciation profiles were evaluated using SHED studies. Running loss, hot soak and diurnal emissions were included in each test. Total VOC emissions measured during each test were compared against MOBILE6 predicted emissions. An evaluation of evaporative emissions impacts on ozone formation potentials using MIR was completed, comparing measured and predicted emissions. The measured:predicted speciation results ranged between 0.93 and 1.11 and had an average value of 1.02. For the conditions tested, MOBILE6 underestimated evaporative emissions in 20 of 24 comparisons. MOBILE6-based ozone formation potentials may be underestimated.