I always dreamed about the machine which tells me to enter long right before the market starts to go up. Might a neural network be this machine? Using Tradesignal and the free Python Neural Net library Pyrenn it is easy to find out… Part one: Classification of data The first step in the process is

Among stock investors, a common strategy/belief held is Value investing — buying stocks that are relative cheaper on price/fundamental ratios. The idea behind why value investing works is that Value stocks are either (1) riskier and/or (2) have been mispriced by the market. In theory, these

Factor investing can be pursued across asset classes Risk premia products sold by investment banks have generated mostly unattractive returns since 2006 The idea of risk premia indices is great, but the implementation has been poor INTRODUCTION Monoculture can be considered the biggest threat to our

I need precise timestamp in order to study the market reaction to news. Sadly, the SEC has not joined the exchanges in providing nanosecond timestamps from GPS-synced rubidium atomic clocks. Rather, it looks like the best EDGAR timestamps I can get from the SEC are only accurate within a couple of

In the prior two posts, investigated: Subspace Projections: feature selection (1/3) Information Geometric: feature selection (2/3) In this post will evaluate feature importance as implemented by Random Forest and compare to Information Geometric approaches. Here is an outline of what would like to

Since my last currency strength candlestick chart post it seemed to make sense to be able to scan said charts for signals, so below is the code for two Octave functions which act as candlestick pattern scanners. The code is fully vectorised and self-contained, and on my machine they can scan more

As mentioned in the prior discussion feature selection (1/3), of primary interest is understanding the contribution of each feature in x⃗ to the outcome or class labeling function f(x⃗ ) . One way to examine this is to understand how the distributions: p(xf) , the probability distribution of

We are proud to announce the release of our newest paper, Rebalance Timing Luck: The (Dumb) Luck of Smart Beta. Abstract Prior research and empirical investment results have shown that portfolio construction choices related to rebalance schedules may have non-trivial impacts on realized performance.

I am often confronted with the problem of trying to reduce a high dimensional feature set to a, smaller, more effective one. Reducing dimension is important for machine learning models as: the volume of the “search space” grows exponentially at a minimum rate of 2d for binary categorical

Digital signal processing (DSP), specifically the use of digital filters, is embedded in many indicators used by technical analysts to study and make trading decisions using time series of stock, bond, currency, commodity, and other financial asset prices. This analysis takes a look at several of

Hedge Funds analyst roles represent some of the most fiercely contested roles in all of Finance (if not any industry). The work is incredibly varied and challenging, making the jobs the long term aim of many ambitious juniors from diverse backgrounds like Computer Science, finance, economics,

The recent rally has left the market short-term overbought by most measures. Short-term overbought often triggers some studies that suggest a downside edge, but when the overbought condition gets very strongly overbought, then those downside edges often disappear. And at some point, rather than

What exactly is “sequence risk?” We’ll get more into the weeds of it, but for now, consider it the risk of loss when you can least afford it. Think of a client leaving their retirement party with their shiny new set of steak knives and then learning via the news that their enormous position in

Risk has always played a very large role in the world of finance with the performance of a large number of investment and trading strategies being dependent on the efficient estimation of underlying market risk. With regards to this, one of the most popular and commonly used representation of risk

As many of you will know by now, Hudson & Thames is pivoting towards an open-core business model and away from our dreams of pure open source and the “unlocking the commons”. What follows is a very brief history of our learnings with open-source. Starting Out MlFinLab started as an ambitious

In financial markets, risk refers to the probability distribution of future returns. Uncertainty is a broader concept that encompasses ambiguity about the parameters of this probability distribution. There are various types of measures seeking to estimate risk and uncertainty: [1] realized and

What is the probability of profit of your next trade? You would think every trader can answer this simple question. Say you look at your historical trades (live or backtest) and count the winners and losers, and come up with a percentage of winning trades, say 60%. Is the probability of profit of

Singapore’s stock market has unique features given its strong sector biases However, despite these, there were no structural factor exposures over time Like in other markets, investors can pursue factor investing to generate outperformance INTRODUCTION One of the stories of how Singapore received

Is Gold a Hedge or Safe Haven Asset during COVID–19 Crisis? Md Akhtaruzzaman (Australian Catholic University), et al. May 15, 2020 The COVID–19 pandemic has shaken the global financial markets. Our study examines the role of gold as a safe haven asset during the different phases of this

This article is accompanied by a Google Colab notebook, which contains all the code and additional mathematical details. You can find the notebook here: https://links.quant-quest.com/KGNotebook What Will You Learn in This Article? In this article we will explore how you can automatically identify

In their paper “Time Series Momentum,” published in the May 2012 issue of the Journal of Financial Economics, Tobias Moskowitz, Yao Hua Ooi and Lasse Pedersen documented significant time-series momentum (trend) in equity index, currency, commodity and bond futures—delivering substantial

It is fairly easy to recognize price momentum with price-based indicators ex-post or with lag. Price based momentum signals tend to have lag issues in recognizing the start and end of a price move as there is a tradeoff between noise and lag [1] that can’t be defeated without future information

Most asset classes are bets on economic growth Diversified endowment-style portfolios are essentially short volatility Long volatility strategies can be used to create anti-fragile portfolios LONG OR SHORT VOLATILITY? In what by now seems like a galaxy far far away, I once worked as an equity

Harry Markowitz’s Modern Portfolio Theory (MPT) was seen as an amazing accomplishment in portfolio optimization, earning him a Nobel Prize for his work. it is based on the hypothesis that investors can optimize their portfolios based on a given level of risk. While this theory works very well

In our last post, we ran through a bunch of weighting scenarios using our returns simulation. This resulted in three million portfolios comprised in part, or total, of four assets: stocks, bonds, gold, and real estate. These simulations relaxed the allocation constraints to allow us to exclude

Density estimation is now a trivial one-liner script in all modern software. What is not so easy is to become comfortable with the result, how well is is my density estimated? we rarely know. One reason is the lack of ground-truth. Density estimation falls under unsupervised learning, we don’t

During past months we made a set of articles analyzing the performance of equity factors and selected systematic strategies during coronavirus crisis. These articles were short-ranged with data only from the start of the year 2020, which is enough for the purpose of the quick blog posts, but very

Following on from the prior post, want to discuss the problem of sample independence. Many machine learning models in finance deal with timeseries data, where samples used in training may be close together in time and not be independent of one another. There are very few features in finance that do

There is a large body of academic research demonstrating that the value premium has been persistent over long periods, pervasive across asset classes (stocks, bonds, commodities, and currencies) and also across and within industries, countries, and regions, robust to various fundamental metrics, and

Interactive Brokers has always been a popular brokerage with systematic traders. Initially this could partially be attributed to the fact that IB provided an Application Programming Interface (API) that allowed quants to obtain market data and place trades directly in code. Many competing brokerages

I think one of the most interesting areas in the data analysis field is Natural Language Processing (NLP). These last years this discipline has grown exponentially and now it’s a huge area with a lot of problems we can attempt to solve, like text classification, translations or text generation In

Recently a student of my CBT course asked why he wasn’t seeing the usual output (including dates) when he selected AmiBroker’s “Detailed Log” option and ran a backtest that utilizes a low-level CBT. The answer is that much of the Detailed Log output comes from AmiBroker’s

The coronavirus crisis reordered many things in economics and finance and you can add asset correlations to the list. After markets crashed in March, followed by a strong (so far) rebound, asset classes have continued to move with an unusually deep and broad degree of unison. High, or at least

I have used machine learning in trading strategies over the past 10 years. However my use of ML has often played a relatively small role in the overall design and success of the strategies. I use ML in specific signals or strategy sub-problems where the data / problem setup tends to have a robust

Our first look at calendar influences was in analyzing the best time during the month to execute dual momentum trades. Studies here, here, and here show that stocks perform best early in the month. This is when institutional investors make changes to their portfolios. Prices then are most

In the search for more and better factors, this article examines the cross-sectional relationship between historical skewness (see Jack’s post here) and the returns on a robust set of assets and documents the premium for taking on skewness risk. The authors construct long/short portfolios across

The alpha of global macro funds has been shrinking consistently over time However, correlations to equities & bonds were low on average, offering diversification benefits Capital allocators have been cautious on the strategy in recent years INTRODUCTION He-Man and the Masters of the Universe was

Nowcasting is a modern approach to monitoring economic conditions in real-time. It makes financial market trading more efficient because economic dynamics drive corporate profits, financial flows and policy decisions, and account for a large part of asset price fluctuations. The main technology

Vitali Apirine, inventor of the OBVM indicator, presented another new tool for the believing technical analyst in the Stocks & Commodities August 2020 issue. His new Compare Price Momentum Oscillator (CPMO) is based on the Price Momentum Oscillator (PMO) by a Carl Swenlin. So we got another

Our last few posts on portfolio construction have simulated various weighting schemes to create a range of possible portfolios. We’ve then chosen portfolios whose average weights yield the type of risk and return we’d like to achieve. However, we’ve noted there is more to portfolio

In this article we begin our discussion of artificial neural networks (ANN). We first motivate the need for a deep learning based approach within quantitative finance. Then we outline one of the most elementary neural networks known as the perceptron. We discuss the architecture of the perceptron

Last week I had the honor of being a guest speaker for the National Association of Active Investment Managers (NAAIM)) webinar series. The topic I discussed was “Quantifiable Edges for Active Investing”. That recording is now available to view on the NAAIM website (email registration required).

Momentum in prices is the tendency of assets that have performed well recently (such as over the prior year) to outperform assets in the same asset class that have performed poorly over the prior year. For a more thorough review of momentum check out this post by Wes Gray. This phenomenon has been

When developing a strategy, exits are often not given a second thought. If you are creating a mean reversion, you may default to using Close greater than the 2-period RSI. If you are trading a trend strategy, you may default to trailing exit using 14-day ATR. You try a bunch of entry filters but

Inside most of the Deep Learning frameworks that are available lies a powerful technique called Automatic Differentiation. If you ever encountered these words but don’t know what they mean or how this procedure works, this post is for you. In a previous post, we saw how to built a deep learning

Technical indicators are the base of algorithmic trading. So wouldn’t it be nice to know tomorrows indicator value in advance? This article is about how to use a simple neural network to do so. Python and Tradesignal will be used to do the programming. A single linear neuron A single neuron /

Some EM asset classes are highly correlated, to the point they can almost be considered interchangeable EM equities and hard-currency government debt are highly correlated to US equities and bonds In crisis times, all EM exposure is sold off and fails to provide meaningful diversification benefits

Can we do impact investing that is both good for us and tastes better? In the past, if an investment had positive non-financial outcomes (positive impact), a return trade-off was expected. Today, some investors find that incorporating aspects such as diversity, stakeholders, and environmental

I just came across an excellent and highly relevant piece of research "A comparison of machine learning model validation schemes for non-stationary time series data" by Matthias Schnaubelt. Features like non-stationarity, concept drift, and structural breaks present serious modelling

The Smart Beta Mirage Shiyang Huang (University of Hong Kong), et al. June 2020 We document sharp performance deterioration of smart beta indexes after the corresponding smart beta ETFs are listed for investments. Adjusted by aggregate market return, the average return of smart beta indexes drops