|
Front Cover |
1 |
|
|
The Science of Algorithmic Trading and Portfolio Management |
4 |
|
|
Copyright Page |
5 |
|
|
Contents |
8 |
|
|
Preface |
16 |
|
|
Acknowledgments |
18 |
|
|
1 Algorithmic Trading |
20 |
|
|
Introduction |
20 |
|
|
Advantages |
22 |
|
|
Disadvantages |
23 |
|
|
Changing Trading Environment |
24 |
|
|
Recent Growth in Algorithmic Trading |
30 |
|
|
Investment Cycle |
34 |
|
|
Classifications of Algorithms |
35 |
|
|
Types of Algorithms |
36 |
|
|
Algorithmic Trading Trends |
39 |
|
|
Trading Venue Classification |
40 |
|
|
Displayed Market |
40 |
|
|
Dark Pool |
40 |
|
|
Grey Pool |
40 |
|
|
Dark Pool Controversies |
41 |
|
|
Types of Orders |
42 |
|
|
Execution Options |
42 |
|
|
The Trading Floor |
44 |
|
|
Research Function |
45 |
|
|
Sales Function |
46 |
|
|
Algorithmic Trading Decisions |
48 |
|
|
Macro-Level Strategies |
48 |
|
|
Step 1—Choose Implementation Benchmark |
49 |
|
|
Step 2—Select Optimal Execution Strategy |
49 |
|
|
Step 3—Specify Adaptation Tactic |
51 |
|
|
Micro-Level Decisions |
52 |
|
|
Limit Order Models |
53 |
|
|
Smart Order Routers |
54 |
|
|
Algorithmic Analysis Tools |
56 |
|
|
Pre-Trade Analysis |
56 |
|
|
Intraday Analysis |
56 |
|
|
Post-Trade Analysis |
57 |
|
|
Rule-Based Trading |
57 |
|
|
Quantitative Techniques |
57 |
|
|
High Frequency Trading |
58 |
|
|
Auto Market Making |
58 |
|
|
Quantitative Trading/Statistical Arbitrage |
60 |
|
|
Rebate/Liquidity Trading |
60 |
|
|
Direct Market Access |
62 |
|
|
Advantages |
63 |
|
|
Disadvantages |
63 |
|
|
2 Market Microstructure |
66 |
|
|
Introduction |
66 |
|
|
Market Microstructure Literature |
68 |
|
|
The New Market Structure |
70 |
|
|
Pricing Models |
75 |
|
|
Order Priority |
76 |
|
|
Equity Exchanges |
76 |
|
|
New NYSE Trading Model |
76 |
|
|
Designated Market Makers |
77 |
|
|
Supplemental Liquidity Providers |
78 |
|
|
Trading Floor Brokers |
79 |
|
|
NASDAQ Select Market Maker Program |
79 |
|
|
Empirical Evidence |
80 |
|
|
Trading Volumes |
80 |
|
|
Market Share |
80 |
|
|
Large and Small Cap Trading |
81 |
|
|
Do Stocks Trade Differently Across the Exchanges and Venues? |
82 |
|
|
Volume Distribution Statistics |
82 |
|
|
Day of Week Effect |
84 |
|
|
Intraday Trading Profiles |
86 |
|
|
Spreads |
86 |
|
|
Volumes |
87 |
|
|
Volatility |
89 |
|
|
Intraday Trading Stability—Coefficient of Variation |
91 |
|
|
Special Event Days |
92 |
|
|
Flash Crash |
95 |
|
|
Empirical Evidence from the Flash Crash |
98 |
|
|
What Should Regulators do to SafeGuard Investors from Potential Future Flash Crashes? |
102 |
|
|
Comparison with Previous Crashes |
103 |
|
|
Conclusion |
104 |
|
|
3 Algorithmic Transaction Cost Analysis |
106 |
|
|
Introduction |
106 |
|
|
What Are Transaction Costs? |
107 |
|
|
What Is Best Execution? |
107 |
|
|
What Is the Goal of Implementation? |
108 |
|
|
Unbundled Transaction Cost Components |
108 |
|
|
1. Commission |
108 |
|
|
2. Fees |
108 |
|
|
3. Taxes |
108 |
|
|
4. Rebates |
109 |
|
|
5. Spreads |
109 |
|
|
6. Delay Cost |
110 |
|
|
7. Price Appreciation |
110 |
|
|
8. Market Impact |
110 |
|
|
9. Timing Risk |
111 |
|
|
10. Opportunity Cost |
111 |
|
|
Transaction Cost Classification |
111 |
|
|
Transaction Cost Categorization |
113 |
|
|
Transaction Cost Analysis |
113 |
|
|
Measuring/Forecasting |
115 |
|
|
Cost versus Profit and Loss |
116 |
|
|
Implementation Shortfall |
116 |
|
|
Complete Execution |
118 |
|
|
Opportunity Cost (Andre Perold) |
119 |
|
|
Expanded Implementation Shortfall (Wayne Wagner) |
120 |
|
|
Implementation Shortfall Formulation |
123 |
|
|
Trading Cost/Arrival Cost |
123 |
|
|
Evaluating Performance |
124 |
|
|
Trading Price Performance |
124 |
|
|
Benchmark Price Performance |
125 |
|
|
VWAP Benchmark |
125 |
|
|
Participation Weighted Price (PWP) Benchmark |
127 |
|
|
Relative Performance Measure (RPM) |
128 |
|
|
Pre-Trade Benchmark |
129 |
|
|
Index Adjusted Performance Metric |
130 |
|
|
Z-Score Evaluation Metric |
131 |
|
|
Market Cost Adjusted Z-Score |
132 |
|
|
Adaptation Tactic |
133 |
|
|
Comparing Algorithms |
134 |
|
|
Non-Parametric Tests |
135 |
|
|
Paired Samples |
136 |
|
|
Sign Test |
136 |
|
|
Wilcoxon Signed Rank Test |
137 |
|
|
Independent Samples |
139 |
|
|
Mann-Whitney U Test |
139 |
|
|
Median Test |
141 |
|
|
Distribution Analysis |
142 |
|
|
Chi-Square Goodness of Fit |
142 |
|
|
Kolmogorov-Smirnov Goodness of Fit |
143 |
|
|
Experimental Design |
144 |
|
|
Proper Statistical Tests |
145 |
|
|
Small Sample Size |
145 |
|
|
Data Ties |
145 |
|
|
Proper Categorization |
146 |
|
|
Balanced Data Sets |
146 |
|
|
Final Note on Post-Trade Analysis |
146 |
|
|
4 Market Impact Models |
148 |
|
|
Introduction |
148 |
|
|
Definition |
148 |
|
|
Example 1: Temporary Market Impact |
149 |
|
|
Example 2: Permanent Market Impact |
149 |
|
|
Graphical Illustrations of Market Impact |
150 |
|
|
Illustration 1—Price Trajectory |
150 |
|
|
Illustration 2—Supply-Demand Equilibrium |
151 |
|
|
Illustration 3—Temporary Impact Decay Function |
154 |
|
|
Example—Temporary Decay Formulation |
156 |
|
|
Illustration 4—Various Market Impact Price Trajectories |
157 |
|
|
Developing a Market Impact Model |
158 |
|
|
Essential Properties of a Market Impact Model |
159 |
|
|
Derivation of Models |
161 |
|
|
Almgren & Chriss—Market Impact Model |
161 |
|
|
Random Walk with Price Drift—Discrete Time Periods |
162 |
|
|
Random Walk with Market Impact (No price drift) |
163 |
|
|
I-Star Market Impact Model |
165 |
|
|
Model Formulation |
166 |
|
|
I-Star: Instantaneous Impact Equation |
166 |
|
|
Market Impact Equation |
167 |
|
|
Derivation of the Model |
167 |
|
|
Cost Allocation Method |
168 |
|
|
I* Formulation |
170 |
|
|
Comparison of Approaches |
172 |
|
|
Underlying Data Set |
173 |
|
|
Imbalance/Order Size |
173 |
|
|
Parameter Estimation Techniques |
176 |
|
|
Technique 1: Two-Step Process |
176 |
|
|
Step 1: Estimate Temporary Impact Parameter |
176 |
|
|
Step 2: Estimate ai Parameters |
177 |
|
|
Technique 2: Guesstimate Technique |
179 |
|
|
Technique 3: Non-Linear Optimization |
179 |
|
|
Model Verification |
179 |
|
|
Model Verification 1: Graphical Illustration |
180 |
|
|
Model Verification 2: Regression Analysis |
180 |
|
|
Model Verification 3: Z-Score Analysis |
180 |
|
|
Model Verification 4: Error Analysis |
181 |
|
|
5 Estimating I-Star Model Parameters |
182 |
|
|
Introduction |
182 |
|
|
Scientific Method |
183 |
|
|
Step 1: Ask a Question |
183 |
|
|
Step 2: Research the Problem |
183 |
|
|
Step 3: Construct the Hypothesis |
183 |
|
|
Step 4: Test the Hypothesis |
183 |
|
|
Step 5: Analyze the Data |
184 |
|
|
Step 6: Conclusion and Communication |
184 |
|
|
Solution Technique |
185 |
|
|
The Question |
185 |
|
|
Research the Problem |
185 |
|
|
Construct the Hypothesis |
190 |
|
|
Test the Hypothesis |
192 |
|
|
Data Definitions |
194 |
|
|
Universe of Stocks |
195 |
|
|
Analysis Period |
195 |
|
|
Time Period |
195 |
|
|
Number of Data Points |
195 |
|
|
Imbalance |
195 |
|
|
Side |
196 |
|
|
Volume |
196 |
|
|
Turnover |
196 |
|
|
VWAP |
197 |
|
|
First Price |
197 |
|
|
Average Daily Volume |
197 |
|
|
Annualized Volatility |
197 |
|
|
Size |
198 |
|
|
POV Rate |
198 |
|
|
Cost |
198 |
|
|
Estimating Model Parameters |
198 |
|
|
Sensitivity Analysis |
200 |
|
|
Cost Curves |
205 |
|
|
Statistical Analysis |
206 |
|
|
Error Analysis |
206 |
|
|
Stock Specific Error Analysis |
208 |
|
|
6 Price Volatility |
212 |
|
|
Introduction |
212 |
|
|
Definitions |
213 |
|
|
Price Returns/Price Change |
213 |
|
|
Average Return |
213 |
|
|
Volatility |
215 |
|
|
Covariance |
215 |
|
|
Correlation |
216 |
|
|
Dispersion |
216 |
|
|
Value-at-Risk |
216 |
|
|
Implied Volatility |
217 |
|
|
Beta |
217 |
|
|
Market Observations—Empirical Findings |
218 |
|
|
Forecasting Stock Volatility |
221 |
|
|
Volatility Models |
221 |
|
|
Price Returns |
222 |
|
|
Data Sample |
222 |
|
|
Historical Moving Average (HMA) |
223 |
|
|
Exponential Weighted Moving Average (EWMA) |
224 |
|
|
Arch Volatility Model |
224 |
|
|
GARCH Volatility Model |
225 |
|
|
HMA-VIX Adjustment Model |
225 |
|
|
Determining Parameters via Maximum Likelihood Estimation |
227 |
|
|
Likelihood Function |
227 |
|
|
Estimation Results |
228 |
|
|
Measuring Model Performance |
228 |
|
|
Root Mean Square Error (RMSE) |
229 |
|
|
Root Mean Z-Score Squared Error (RMZSE) |
229 |
|
|
Outlier Analysis |
230 |
|
|
Results |
230 |
|
|
Problems Resulting from Relying on Historical Market Data for Covariance Calculations |
233 |
|
|
False Relationships |
233 |
|
|
Degrees of Freedom |
238 |
|
|
Factor Models |
240 |
|
|
Matrix Notation |
242 |
|
|
Constructing Factor Independence |
243 |
|
|
Estimating Covariance Using a Factor Model |
244 |
|
|
Types of Factor Models |
246 |
|
|
Multi-Index Models |
247 |
|
|
Macroeconomic Factor Models |
247 |
|
|
Cross-Sectional Multi-Factor Models |
248 |
|
|
Index Model |
246 |
|
|
Single Index Model |
246 |
|
|
Statistical Factor Models |
250 |
|
|
7 Advanced Algorithmic Forecasting Techniques |
254 |
|
|
Introduction |
254 |
|
|
Trading Cost Equations |
255 |
|
|
Model Inputs |
255 |
|
|
Trading Strategy |
256 |
|
|
Percentage of Volume |
256 |
|
|
Trading Rate |
257 |
|
|
Trade Schedule |
257 |
|
|
Comparison of POV rate to Trade Rate |
258 |
|
|
Trading Time |
258 |
|
|
Trading Risk Components |
259 |
|
|
Trading Cost Models—Reformulated |
260 |
|
|
Market Impact Expression |
260 |
|
|
I-Star |
260 |
|
|
Market Impact for a Single Stock Order |
260 |
|
|
Market Impact for a Basket of Stock |
262 |
|
|
Timing Risk Equation |
262 |
|
|
Timing Risk for a Basket of Stock |
267 |
|
|
Comparison of Market Impact Estimates |
267 |
|
|
Volume Forecasting Techniques |
270 |
|
|
Daily Volumes |
270 |
|
|
Definitions |
270 |
|
|
Daily Forecasting Analysis—Methodology |
271 |
|
|
Variable Notation |
271 |
|
|
ARMA Daily Forecasting Model |
271 |
|
|
Analysis Goal |
272 |
|
|
Forecast Improvements |
276 |
|
|
Daily Volume Forecasting Model |
276 |
|
|
Forecasting Monthly Volumes |
277 |
|
|
Forecasting Covariance |
282 |
|
|
Efficient Trading Frontier |
284 |
|
|
Single Stock Trade Cost Objective Function |
286 |
|
|
Portfolio Trade Cost Objective Function |
286 |
|
|
8 Algorithmic Decision Making Framework |
288 |
|
|
Introduction |
288 |
|
|
Equations |
289 |
|
|
Algorithmic Decision Making Framework |
291 |
|
|
1) Select Benchmark Price |
291 |
|
|
Arrival Price Benchmark |
291 |
|
|
Historical Price Benchmark |
292 |
|
|
Future Price Benchmark |
294 |
|
|
Comparison of Benchmark Prices |
295 |
|
|
2) Specify Trading Goal |
295 |
|
|
1. Minimize Cost |
296 |
|
|
2. Minimize Cost with Risk Constraint |
298 |
|
|
3. Minimize Risk with Cost Constraint |
299 |
|
|
4. Balance Trade-off between Cost and Risk |
299 |
|
|
5. Price Improvement |
300 |
|
|
Further Insight |
302 |
|
|
3) Specify Adaptation Tactic |
303 |
|
|
Projected Cost |
304 |
|
|
Target Cost Tactic |
307 |
|
|
Aggressive-in-the-Money |
308 |
|
|
Passive-in-the-Money |
310 |
|
|
Comparison across Adaptation Tactics |
312 |
|
|
Modified Adaptation Tactics |
313 |
|
|
How Often Should We Re-Optimize Our Tactics? |
313 |
|
|
9 Portfolio Algorithms |
316 |
|
|
Introduction |
316 |
|
|
Trader’s Dilemma |
317 |
|
|
Variables |
318 |
|
|
Transaction Cost Equations |
319 |
|
|
Market Impact |
320 |
|
|
Price Appreciation |
320 |
|
|
Timing Risk |
321 |
|
|
One-Sided Optimization Problem |
321 |
|
|
Optimization Formulation |
321 |
|
|
Constraint Description |
322 |
|
|
Objective Function Difficulty |
324 |
|
|
Optimization Objective Function Simplification |
324 |
|
|
Portfolio Optimization Techniques |
325 |
|
|
Quadratic Programming Approach |
325 |
|
|
Trade Schedule Exponential |
327 |
|
|
Residual Schedule Exponential |
328 |
|
|
Trading Rate Parameter |
329 |
|
|
Market Impact Expression |
329 |
|
|
Timing Risk Expression |
330 |
|
|
Comparison of Optimization Techniques |
331 |
|
|
Portfolio Adaptation Tactics |
335 |
|
|
Description of AIM and PIM for Portfolio Trading |
336 |
|
|
How Often Should We Re-Optimize? |
338 |
|
|
Managing Portfolio Risk |
339 |
|
|
Residual Risk Curve |
339 |
|
|
Minimum Trading Risk Quantity |
341 |
|
|
Maximum Trading Opportunity |
342 |
|
|
When to Use These Values? |
343 |
|
|
Program-Block Decomposition |
344 |
|
|
Appendix |
347 |
|
|
10 Portfolio Construction |
350 |
|
|
Introduction |
350 |
|
|
Portfolio Optimization and Constraints |
351 |
|
|
Transaction Costs in Portfolio Optimization |
354 |
|
|
Portfolio Management Process |
358 |
|
|
Example: Efficient Trading Frontier w/ and w/o Short Positions |
359 |
|
|
Example: Maximizing Investor Utility |
359 |
|
|
Trading Decision Process |
360 |
|
|
Unifying the Investment and Trading Theories |
362 |
|
|
Cost-Adjusted Frontier |
367 |
|
|
Determining the Appropriate Level of Risk Aversion |
369 |
|
|
Best Execution Frontier |
370 |
|
|
Portfolio Construction with Transaction Costs |
371 |
|
|
Quest for best execution frontier |
373 |
|
|
Return |
374 |
|
|
Risk |
374 |
|
|
Conclusion |
378 |
|
|
11 Quantitative Portfolio Management Techniques |
380 |
|
|
Introduction |
380 |
|
|
Are the Existing Models Useful Enough for Portfolio Construction? |
382 |
|
|
Current State of Vendor Market Impact Models |
383 |
|
|
Pre-Trade of Pre-Trades |
386 |
|
|
Estimation Process |
387 |
|
|
Applications |
391 |
|
|
Example 1 |
391 |
|
|
Example 2 |
392 |
|
|
Example 3 |
392 |
|
|
Example 4 |
392 |
|
|
How Expensive Is It to Trade? |
393 |
|
|
Acquisition and Liquidation Costs |
396 |
|
|
Portfolio Management—Screening Techniques |
399 |
|
|
MI Factor Scores |
403 |
|
|
Derivation of the MI Factor Score for Shares |
403 |
|
|
Current State of MI Factor Scores |
405 |
|
|
MI Factor Score Analysis |
405 |
|
|
Alpha Capture Program |
407 |
|
|
Example 5 |
408 |
|
|
Example 6 |
409 |
|
|
Alpha Capture Curves |
412 |
|
|
12 Cost Index & Multi-Asset Trading Costs |
414 |
|
|
Introduction |
414 |
|
|
Cost Index |
415 |
|
|
Cost Basis |
416 |
|
|
Cost Strategy |
417 |
|
|
Normalization Process |
419 |
|
|
Customized Indexes |
421 |
|
|
Real-Time Cost Index |
422 |
|
|
Back-Testing |
427 |
|
|
Market Impact Simulation |
429 |
|
|
Simulation Scenario |
431 |
|
|
Multi-Asset Class Investing |
434 |
|
|
Investing in Beta Exposure and Other Factors |
434 |
|
|
Beta Investment Allocation |
438 |
|
|
Multi-Asset Trading Costs |
439 |
|
|
Global Equity Markets |
440 |
|
|
Multi-Asset Classes |
441 |
|
|
13 High Frequency Trading and Black Box Models |
448 |
|
|
Introduction |
448 |
|
|
Data and Research |
450 |
|
|
Strategies |
451 |
|
|
Statistical Arbitrage |
451 |
|
|
Triangular Arbitrage |
455 |
|
|
Liquidity Trading |
458 |
|
|
Market-Neutral Arbitrage |
459 |
|
|
Index and Exchange Traded Fund Arbitrage |
461 |
|
|
Merger Arbitrage |
462 |
|
|
Evaluation |
465 |
|
|
Summary |
469 |
|
|
References |
472 |
|
|
Index |
484 |
|