Algorithmic Trading Strategies [Nick Firoozye]

Algorithmic Trading Strategies
Nick Firoozye​

Пример лекции 15 "Momentum - a first glance. We introduce the very basic intution behind momentum and how we would construct the most simplistic of strategies"


Этот курс по алгоритмической торговле охватывает основополагающие принципы алгоритмической торговли, включая стратегии следования за трендом (trend-following), carry трейдинг, value трейдинг, возврата к среднему значению и относительной стоимости (mean-reversion and relative value strategies), а также других более сложных стратегий, таких как короткая гамма (short-gamma). Мы обсудим обоснование стратегий, проекты стандартных стратегий, плюсы и минусы различных вариантов дизайна стратегий и ожидаемые выгоды от диверсификации в портфельных стратегиях. Наконец, поскольку алготрейдинг подвержен переобучению (overfitting) и, как следствие, показывает низкую выгоду (poor performance), мы обсудим p-hacking ('financial charlatanism') и различные стратегии, позволяющие этого избежать. Курс основан на математическом и статистическом обосновании и доказательстве, а также содержит описание и свойства каждой стратегии.

После этого курса вы сможете анализировать, разрабатывать и подтверждать количественные торговые стратегии (quantitative trading strategies):

- Научитесь понимать механику стандартных реализаций торговых стратегий, основанных на рисках и премиях для одного актива и портфеля.
- Научитесь определять "за" и "против" различных подходов к разработке стратегий и распространенных ошибок, с которыми сталкиваются алгоритмические трейдеры.
- Сможете разрабатывать новые и улучшать алгоритмические стратегии.
- Научитесь понимать, когда часто используемые стратегии работают, а когда нет.
- Научитесь определять статистические свойства стратегий и различать математически доказанные стратегии от эмпирических.
- Курс дает понимание методов, чтобы предотвратить переобучение.
- Ознакомитесь с широкой областью алгоритмических торговых стратегий.
- Овладеете основополагающей теорией и механикой самых распространенных стратегий.
- Получите понимание принципов и контекста, необходимых для новых научных исследований большого количества открытых вопросов в этой области.

Автор курса:

Доктор Ник Фирозью (Nick Firoozye) - ученый и статистик с более чем 20-летним опытом работы в финансовой индустрии, как в компаниях, занимающихся покупкой, так и продажей (buy and sell-side firms), в основном в сфере исследований. Он начал свою карьеру в Lehman Brothers, занимаясь моделированием MBS/ABS, возглавляя команды по стратегии портфеля и исследованиям EM Quant, а затем занимал различные руководящие должности в Goldman Sachs и DeutscheBank, а также в управляющих активами Sanford Bernstein и Citadel.
В настоящее время он является управляющим директором и главой Global Derivative Strategy, входящей в группу количественных стратегий, в Nomura, а также является почетным старшим преподавателем по компьютерным наукам в University College London (UCL), специализируясь на Robust Machine Learning в области финансов.
Доктор Ник Фирозью - соавтор книги "Управление неопределенностью, снижение риска - Борьба с неизвестным в оценке финансовых рисков и принятии решений" ("Managing Uncertainty, Mitigating Risk - Tackling the Unknown in Financial Risk Assessment and Decision Making") о роли неопределенности и неточной вероятности в финансах в свете многих недавних финансовых кризисов.
Сейчас Доктор Ник Фирозью пишет книгу об алгоритмических торговых стратегиях, основанных на его научной работе по той же теме в UCL, и на данном онлайн-курсе в Experfy.

Курс на английском.

Продолжительность: 62 лекции, 6 часов.

Спойлер: Описание и содержание курса на английском

Duration: 62 lectures, 6h

Course Description

This algorithmic trading course covers the underlying principles behind algorithmic trading, including analyses of trend-following, carry, value, mean-reversion, and relative value strategies and other more obscure strategies like short-gamma. We will discuss the rationale for the strategy, standard strategy designs, the pros and cons of various design choices, and the gains from diversification in portfolio strategies. Finally, since the industry is plagued by overfitting and resulting poor performance, we will discuss p-hacking (or 'financial charlatanism') and various strategies to avoid it. We are focusing on the mathematical and statistical justification, formulation and properties of each strategy.

After this course you will able to Analyse, Design and Confirmate quantitative trading strategies:

- Understand the mechanics of standard implementations of the single asset and portfolio based risk-premia trading strategies.
- Recognize pros and cons of various approaches to designing strategies and the common pitfalls encountered by algorithmic traders.
- Be able to devise new and improve algorithmic strategies.
- Recognize the reasons commonly-used strategies work and when they don't.
- Understand the statistical properties of strategies and discern the mathematically proven from the empirical.
- Acquire an understanding of methods to prevent overfitting.
- Gain familiarity with the broad area of algorithmic trading strategies.
- Master the underlying theory and mechanics behind the most common strategies.
- Acquire the understanding of principals and context necessary for new academic research into the large number of open questions in the area.

Instructor: Nick Firoozye
Dr. Nick Firoozye is a data scientist & statistician with over 20 years of experience in the finance industry, in both buy and sell-side firms, largely in research. He started his career in Lehman Brothers doing MBS/ABS modeling, heading teams in portfolio strategy and EM quant research, later taking a variety of senior roles at Goldman Sachs, and DeutscheBank, and at the asset managers, Sanford Bernstein, and Citadel. He is currently Managing Director and Head of Global Derivative Strategy, part of the Quantitative Strategy Group, at Nomura. He is currently an Honorary Senior Lecturer in Computer Science at University College London (UCL), focusing on Robust Machine Learning in finance. He recently co-authored a book, entitled «Managing Uncertainty, Mitigating Risk» about the role of uncertainty and imprecise probability in finance, in light of the many recent financial crises, and he is writing a book on Algorithmic Trading Strategies based on his recent Ph.D course on the same topic offered at UCL and current online course at Experfy.

Curriculum

Module 1: Course Overview 28:14

Lecture 1 Overview 05:18
We discuss algo trading strategies and their recent context in the world of alternative investment management.

Lecture 2 Context and Background 09:10
Introduction to the area, Algo as opposed to High-Frequency/Low Latency Trading, and areas of growth. The goals of the course, for students/academics, professionals, and algo traders, and general background to the course.

Lecture 3 What the course is Not and the Role of Data Science 06:18
What the course is not. The Role of Data science and ML - do data scientists need to know about 'canonical' strategies? Can they just start fresh? We argue that some of the most commonly used strategies give good guidance for data scientists whose techniques rarely work "out of the box" and are especially prone to problems in the area of algo trading strategies.

Lecture 4 Prerequisites and Syllabus 03:34

Lecture 5 Syllabus 03:54
We describe the basics of the syllabus. Some of these materials are covered very thoroughly, while others are covered quite quickly as methods in use / approaches to consider in devising and refining strategies. We cover Background, Momentum, Mean Reversion, Carry, Value, Basic Portfolio Strategies, and the important concept of Overfitting, focusing on the mathematical and statistical justification, formulation and properties of each strategy.

Resource 1 - Slides on Introduction, Background Material, Goals and Prerequisites and Syllabus.

Module 2: Industry Overview and Math Review 56:52

Lecture 6 Industry Overview 05:10
Alternatives, Hedge Funds, CTAs and Quant Funds. What size and what numbers? How much are they growing? Where are the opportunities? From the top down look at the overall prospects of the industry where Algo Trading Strategies are employed.

Lecture 7 Tracking Funds 05:42

Lecture 8 Tracking Benchmarks 04:36

Lecture 9 Styles 04:39

Lecture 10 Algo Trading Strategy Infrastructure 04:38

Quiz 1 Intro Quiz on Background
Quiz on background and introduction

Resource 2 - PDF Slides

Lecture 11 Review 2 ARMA Processes 09:46
We review basic Box-Jenkins method for ARMA models, look at characteristic polynomials, describe stationary vs nonstationary processes

Lecture 12 Review 1 - White noise and Brownian motion 07:12
We review some of the basic mathematics for timeseries including white noise and brownian motion

Lecture 13 Review 3 - Autocovariance, autocorrelation and criteria 05:00
We review the ACF and its relation to ARMA models, and start on criteria (AIC, BIC) as a means of doing model choice.

Lecture 14 Review 4 - Cross Validation, Bootstrap and solving SDEs 10:09
We touch on more computer intensive methods for doing model selection - cross validation and finding standard errors-bootstrap. Finally, we discuss two most common method for solving SDEs in closed-form, muitipying constants and integration by parts/Ito's lemma

Quiz 2 Basic ARMA models
We go through some basic ARMA models and their ACFs

Module 3: Momentum / Trend Following 01:16:53

Lecture 15 Momentum - a first glance 04:26
We introduce the very basic intution behind momentum and how we would construct the most simplistic of strategies

Lecture 16 Momentum Related Factoids 1 06:16
We discuss some of the properties and tradeoffs of momentum, many of which can be changed by strategy design.

Lecture 17 Momentum Factoids 2 05:01
Further factoids including examples of returns in practice

Lecture 18 Proving results about momentum 1 07:02
We look at discrete time versions of momentum and seek to prove that skewness changes by horizon

Lecture 19 Proving results about momentum 2 10:08
This is a whiteboard section on the basics of the skewness over horizon results (Martin-Zou), going through the proof, showing that the concepts are relatively easy (even if the algebra is a little tedious).

Lecture 20 Skewness - why is it so strange? 03:58
Having proved results about the skewness of momentum returns over different horizons, we apply it to an exponentially weighted moving average (EWMA) rule, showing how the peak skewness is related to the effective lookback (in our case, the "span") of the EWMA.

Lecture 21 Practical Momentum - Different methods for similar results 07:21
We describe the most commonly used methods in the industry, from Kalman Filters to Moving Averages to ARIMA models. Used properly, most of these models can attain almost the same performance.

Lecture 22 Coding Momentum 1 10:46
We introduce an ipython notebook. It takes data from Quandl (and some from Yahoo finance) including SPX, SPTR, and Effective Fed Funds. We use these to construct S&P 500 excess returns, and compare to SPX. We then devise a strategy for momentum.

Lecture 23 Coding Momentum 2 05:38
Computing relevant stats (Sharpes and Skewness) over different horizons

Lecture 24 Momentum variants, and fads and fancies in models 05:10
Cross sectional vs Timeseries momentum. Which is better? Where are each of them used? Why should we know them both? Fads and fancies in momentum modelling. Models vs Method.

Lecture 25 Momentum - capped, floored and otherwise altered signals 03:45
We look at Winsorising or capping and flooring the signals (sometimes needed to prevent too large capacity utilisation), using thresholds, etc. These typically detract from the skewness, but they could help the overall performance. We look at various methods and discuss their pros and cons and how to measure them.

Lecture 26 Readings for further study 04:20
We give links to and summarize the handful of most important papers on statistical aspects of momentum trading for further study. Being well-known, these are also the most cited papers, and so any new academic research can be found (using google scholar) just by searching preprints and papers which cite these important studies.

Lecture 27 Momentum – Summary 03:02
Summarizing the main points we made in section 2 on Momentum

Module 4: Mean Reversion / Change-points 01:46:02

Lecture 28 Mean Reversion Overview and Time-scales of trades 08:03
Overview of MR, and the timescales/horizons associated with MR, Momentum and Value

Lecture 29 Putting timescales all together and where to search for history 06:41
A continuation of the previous lecture, putting the timescales all together, and looking to ancient history (if need be)

Lecture 30 Mean Reversion in action 04:28
The typical features of an MR trading strategy, what to expect and what to be careful with

Lecture 31 Rationales for Mean Reversion 06:56
Various competing (or not so competing) rationales for mean reversion: Liquidity Provision and Overreaction

Lecture 32 Vol and Mean Reversion 07:46
Volatility and Mean Reversion, the theory and empirics behind their relationship

Lecture 33 Liquidity – References 03:05
A few of the most important academic papers on liquidity

Lecture 34 Mean Reversion and Unit Root Tests, Intro 04:51
An analysis of the types of behaviour we want to discern between, focusing on mean reverting vs unit root processes.

Lecture 35 Augmented Dickey Fuller Tests 05:19
ADF Tests are the most commonly used unit root tests out there. We introduce their use and limitations

Lecture 36 KPSS Tests 03:45
KPSS tests turn H0 and H1 on their heads, testing for mean-reversion. They also have their limitations

Lecture 37 Variance Ratio Tests 04:08
We introduce variance ratio tests, explore their use and misuses

Lecture 38 Cointegration and Johansen Test 09:35
Cointegration and Engle Granger testing, and the more thorough Johansen test

Lecture 39 Harvey Nyblom Tests and Shortcomings 04:27
Harvey Nyblom is to Johansen as KPSS is to ADF and we explore H-N Tests and then the shortcomings for all testing methods

Lecture 40 Power, Type I and Type II errors 04:16
power of tests, confidence intervals, type 1 and type 2 errors

Lecture 41 RV Trades 04:47
RV Trade ideas and MR

Lecture 42 Filters 07:24

Lecture 43 Changepoints - Overview 08:57
Overview and more classical approaches to changepoint detection. These are useful for piecewise linear fits to data to establish trending means and mean reversion to these trending means.

Lecture 44 Changepoints - Lasso based tools 06:17
Using the lasso regression to detect trends, we can identify breakpoints and extract trends at the same time. While not always the easiest method, regularisation methods like lasso are helpful in many circumstances and also are a decent framework to think of the underlying problems.

Lecture 45 Changepoints - sequential binary segmentation, switching kalman filters and summary 05:17
We follow up with a very practical and implementable tool - sequential binary segmentation (and Wild binary segmentation)

Resource 3 - Quiz 3

Module 5: Carry, Value, and Portfolio Strategies 54:24

Lecture 46 Carry - First definitions 06:05
We define carry and give a rationale in terms of P vs Q measures

Lecture 47 P vs Q measure 04:23
We continue the discussion of the differences between P measure (physical world) vs Q measure (for pricing and hedging derivatives). While Q (where spot rates will always drift towards forwards or - 'forwards are realised') is an interesting construct, it is merely that. We have to use it to price and hedge (or 'risk manage') derivatives. Realistically, in incomplete markets, Q is not actually unique and is merely a useful construct. Realistically speaking, spot rates tend to stay put, and random walks are much more likely than having realised forwards. If spot rates are martingales/random walks, this is a perfectly decent rationale for studying carry.

Lecture 48 Defining Carry 03:52
Defining carry - what is it? Why do we care about it? What is a positive carry position and what is a negative carry position? What about commodities?

Lecture 49 Carry for Swaps (and a little for bonds) 05:03
We define carry for swaps, something not as easily available, and also a little bit for bonds. Bonds, however, are altogether more difficult, since you need to know bond-specific funding rates (term repo rates), so we mostly pursue carry for swaps.

Lecture 50 Carry for Futures, FX, Equities and Derivatives 06:50
We briefly describe carry for Futures (including commodity and equity) and FX and for the less well covered area of Derivatives.

Lecture 51 Carry - Summary 03:31
We summarize the exploration of carry

Lecture 52 Value 06:16
We define value, its use and how it differs from Equities (where it is well defined and followed regularly) to fixed income, fx and commodities. Value, with its longer-term mean-reversion properties, is naturally orthogonal to momentum, and mean-reversion.

Lecture 53 Portfolio Strategies 1 - MVO 06:43
Mean variance optimisation as a guide to basics of portfolio strategy

Lecture 54 Portfolios - Testing weights 05:07
We present portfolio optimisation as a regression and describe F-tests for statistical significance of changes in portfolio weights.

Lecture 55 Portfolio Optimisation - Conditional Portfolios and other performance measures 06:34
We introduce conditional portfolios and optimisation to include dynamic reallocation. Using augmented portfolios allows us to consider dynamic signals in portfolio optimisation. Finally, we talk about the shortcomings of most MVO style portfolio optimisation, and introduce a number of the standard performance measures used in measurement and allocation problems.

Resource 4 Slides as PDF

Module 6: Overfitting 36:41

Lecture 56 Intro to Overfitting and the major issues 04:16
We introduce the problem and related issues of p-hacking, lack of reproducibility, and holdout overfitting in Kaggle competitions.

Lecture 57 Overfitting in Finance 05:27
Overfitting in finance is perhaps more problematic than any other field. While Amazon or Google could miss a few keyclicks by relying on spurious results, in finance, we could easily risk insolvency. Meanwhile, overfitting is altogether too common and recent studies have shown its prevalence.

Lecture 58 Dealing with overfitting - increasing backtest length 04:02
Bailey et al have proposed increasing backtest lengths to avoid overfitting. The method is illustrative but provides more of a rule of thumb. We describe the results of their paper on "Financial Charlatanism and Pseudo-Mathematics" and the concept of minimum backtest length

Lecture 59 Adjusted Sharpe Ratios and Multiple Hypothesis Tests 06:11
Harvey and Liu discuss the statistics of Sharpe ratios, converting to p-values (if Sharpe = E[Ret]/Std[Ret], the test is H0: E[Ret]=0). They then discuss multiple hypothesis testing and how one deals with it.

Lecture 60 Multiple Hypothesis Testing - Holm and Bonferroni 06:55
Ways of dealing with Multiple Hypothesis Testing - Holm and Bonferroni methods, somewhat more extreme than optimal but giving some good insight into means of adjusting p-values.

Lecture 61 Multiple Hypothesis Testing - BHY adjustments and Practical Methods to prevent overfitting 09:50
We describe the best method for controlling the rate of false discovery (FDR), the BHY adjustment and we talk about its impact on Sharpe Ratios based on number of strategies run and size of history available for backtest. Finally, we summarize the practical approaches to backtest overfitting.

Resource 5

Module 7: Course Summary 07:08

Lecture 62 Course Summary 07:08