NilssonHedge asked OpenAI to write a short success story for one of our many users. It turned out to be pretty ok. We do apologize in advance for the somewhat chatty nature of this post.
The databases are updated and available for download.
Once upon a time, in a city of finance and business, there was a young man named Jack who was trying to make it big in the world of investing. He had a natural talent for numbers and was always looking for new ways to maximize his returns.
One day, while searching for information on the latest hedge fund trends, Jack stumbled upon NilssonHedge, a database that collected returns and other important information on various hedge funds. He was fascinated by the wealth of available information and quickly became a frequent visitor to the site.
As he delved deeper into the data, Jack realized that there were patterns and trends emerging. He could identify which hedge funds were performing the best and which were struggling. He also discovered that the funds with the highest returns often had certain characteristics in common, such as a focus on trend strategies.
Excited by this new knowledge, Jack decided to start his own fund of hedge funds, and he used NilssonHedge as a valuable resource in building and managing his portfolio. He was able to make informed decisions based on the information he had access to and his fund quickly became one of the top performers in the market.
Word of Jack’s success spread quickly, and soon, other investors and fund managers were seeking out NilssonHedge to help guide their investment decisions. The database became the go-to resource for anyone looking to stay ahead of the curve in the hedge fund industry.
And so, NilssonHedge went from being a small and relatively unknown database to one of the most trusted and respected sources of information in the world of finance. Jack’s success was a testament to the power of data and the impact that NilssonHedge had on the industry.
From that day forward, NilssonHedge remained the standard by which all other hedge fund databases were judged, and its influence continued to grow and shape the world of finance for years to come.
We have been working behind the scenes the last few weeks and have created additional refinements when it comes to diving deeper into benchmarking exercises. We have created the following indices:
While a lot of the indices are correlated (e.g. CTA Momentum and the Systematic CTA index), the drifts will be slightly different and may be useful for benchmarking style pure managers or understanding specific parts of the market structure.
We are touching up the final classification of the underlying managers and may revise the constituents list in the next few days.
We are also working on adding a “Volatility Arbitrage” index which should be ready in a few weeks.
The Quant Index has been defined as managers that are engaged in primarily non-trend strategies (AI, Fundamental, Short Term, etc.) and has generally delivered an interesting risk profile. The Discretionary Commodity index is the least correlated to the broader CTA Indices with a correlation of 0.2.
In particular, for investors looking for diversification, the CTA Quant Index has a relatively low correlation to the general CTA index of 0.67. Momentum on the other hand is 0.95 correlated to our broader CTA index.
In other news, we have distributed the first 2023 version of the NilssonHedge database and given the risk-on nature of the month, traditional alternative strategies did really well. Crypto Managers generated a whopping 32% highlighting their Cathie Wood / ARK correlation.
We start the year with some 1,267 strategies with data, representing 67% of the strategies in the database.
On the CTA side, consisting of a large share of private vehicles 222 managers have reported numbers as of “pixel-time”.
If you are interested to explore the structure of our new indices, please head over to our shop and subscribe to the free index data.
During the last few weeks, we have been engaged in a number of housekeeping issues. One of them is our index reconstitution. We will be launching several new CTA indices that give additional details for specific sectors.
Every year we reconstitute our indices, a largely automated process, but which involves a fair amount of discretionary oversight to ensure that our indices are representative and relevant. This year, we have some 2900 different strategies in our various composites. Not all managers fit nicely into one composite.
The indices are disseminated through our database, website, partners, and the analytical software packages that NilssonHedge has partnered up with. These indices are generally not proper benchmarks from a regulatory perspective but rather performance indicators.
As a small notice, these indices are not designed to be investable and represent the average return from managers which requires continuous rebalancing. Index figures are updated continuously throughout the month and reflect the current state of the database.
Except for carveouts, there will be no backtracking of the indices. In cases when a new index is launched, it will be based on the historical, actual state of the database as of the launch date. Carveouts (i.e. a separate benchmark based on an existing index; CTA Systematic or CTA Discretionary managers are two examples).
Return data will be continuously updated and will incorporate lagging managers who for various reasons have decided to restart reporting. This may introduce variability in the historical rate of returns.
Remember that NilssonHedge calculates strategy composites based on available data and our individual return stream may not perfectly match the managers’ official track records. We seek to give researchers and allocators a better view of the average investor experience, rather than the asset-weighted investor experience that would be the typical outcome from traditional track records. This may cause minor deviations for particular managers.
Constituents
Every index starts with the index members. Members are selected based on the semi-systematic framework, based on the NilssonHedge classification framework. Below we describe the algorithm for selecting strategies that will form part of an index.
Our methodology for selecting constituents is straightforward and based on the following rules:
At the end of January, each year, NilssonHedge isolates the managers and strategies that have reported returns for December, in the prior year. In this case, we need data for December 2022 at the end of this month (January 2023). We put emphasis on timely reporting and disclosure of results.
This list remains fixed throughout the year, regardless of the number of new managers in the database. As such, expect the number of managers to gradually decline throughout the year.
Given the indices are calculated based on the average, the index is rebalancing toward reporting managers on a monthly basis.
There are currently no requirements on assets under management, length of track record, or other qualitative aspects of the underlying firms.
The monthly indices are generally not backfilled and represent the out-of-sample results. The exception is if we create a carveout of a particular subindex.
For instance, our Discretionary CTA index was launched in 2020, but historical returns were based on the structure and manager classification of the database as of December 2018.
In such cases, the indices will include managers that have stopped reporting to avoid survivorship biases.
The daily indices have been partially backfilled based on the status of the database. We disclose the “go-live” date for affected indices.
NilssonHedge is discretionarily classifying strategies based on a semi-automated assessment of strategies, based on among other things keywords and potentially additional touch points with added managers. Our keyword search algorithm is similar to the one described in “Man vs. Machine: Comparing Discretionary and Systematic Hedge Fund Performance”.
Classification Framework
Some managers are easy to classify, and some other managers are not easy to classify due to their hybrid investment strategy. The classification always involves a certain amount of discretionary judgment.
We classify managers into four different categories:
Type: The headline strategy, to name a few: CTA, Crypto, Market Neutral. We may not follow the managers’ self-classification of their own strategy but try to classify them based on our perception of the strategy and how the return stream behaves in terms of volatility, correlation, and betas to markets.
As an example, is that our CTA encompasses classical Managed Futures managers, Global Macro managers, and asset managers that may not have registered with the NFA and may thus not formally hold themselves out to be a “CTA”.
Likewise, we do not have a pure classification for Macro Managers, that may engage in trading of cash securities, swaps, and other instruments that are not typically part of the classical toolbox for a CTA.
Style: Here we typically classify a manager depending on their investment framework. The main division is if the manager is systematic or discretionary.
As practitioners may have observed even strictly systematic managers have a large degree of discretionary elements, and discretionary traders may use systematic elements or models.
Strategy: Based on the specific investment strategy implemented we further classify the return stream into a number of sub-strategies.
Trend Following, Fundamental, and AI are sample strategies. The Strategy category is more fluent.
Super strategies exist:
Quant, a new index, which consists of AI, Behavioral, Contrarian, Fundamental, Market Neutral, Short Term subject to a systematic implementation.
Sector: The final classification is based on the sectors a program has exposure to.
Super Sector:
Commodities consist of all commodity sectors’ specific managers (Energy, Metal, and Agricultural) and managers that may have exposure to all commodity markets without a disclosed focus.
Transparency
Each week, we update the performance of our indices and present them in the below format via the NilssonHedge webpage. Moreover, we disclose the index members, per year. This can be found in a tabular form at the bottom for each Nilsson Hedge index.
Login
