AiLA Indices is a Benchmark Administrator headquartered in Singapore. We offer two types of Products: Alpha Products to generate absolute returns for investors and Tracker Products which are used by our clients to replicate returns of an illiquid asset using a basket of liquid assets.
The alpha is generated from three parts a) when to invest b) how much to invest c) in what asset to invest. These three decisions are dependant every day on the market behaviour which includes allocations of investors in the constituent assets. Machine Learning addresses questions a & c, whereas the daily rebalancing methodology addresses question b. At the end of each day after market close, the systematic processes decides on direction, quantum and tenor of investment and here the machine learning provides the edge.
Consider the example of H/J Spread in RBOB, the RVP changes affect this spread and becomes a leading indicator of driving demand in summer and that in turn is reflected in the M-Z spread. This information is captured, normalized and taught to the system to decide on direction of allocation in RBOB asset. More details are available on our perspective page https://ailaindices.com/perspectives.php.
The live performance is running on a set a model with a fixed set of hyperparameters. This performance is out of sample and is recorded and calculated by an independent calculation agent. For more details on how we address overfitting, please refer to our research piece https://ailaindices.com/analysing-overfitting.php.
AiLA Products are by design opportunistic and can be market neutral. Most standard commodity indices are long only and constantly invested. AiLA provides allocation flexibility in both direction and timing.
Yes, AiLA Products are replicable with a large computing environment and are not designed for an xl sheet. AiLA Products are IOSCO compliant and follows a rigorous approach in development of Products.
While AiLA might also use a factor driven approach, we focus on the timing of allocation. We train our models to recognise favourable risk/return opportunities and to make a trade when such opportunities arise. Similarly, we exit positions when these opportunities no longer exist. This methodology results in daily rebalancing and these dynamic weights generates a differentiated set of returns. More details can be found on https://ailaindices.com/aila-investment-strategies.php.
AiLA provides its products to a variety of clients. This includes i) Investment Banks who provide swaps on AiLA strategies to their Asset Management clients, ii) ETF Issuers who launch ETFs tracking AiLA product, iii) Hedge funds who source our Products directly and iv) Trading desks of Corporates who invest in our strategies.
We have an independent buy and sell model for each asset used in an product. We define a maturity of a commodity as an asset. As an example, CLZ (December WTI ) is an asset. Therefore, a given product consisting of multiple constituent assets could be driven by a large number of underlying models. This diversified approach gives both independent and uncorrelated returns.
AiLA uses a mix of Macro and Micro Factors to train each model, with each asset having a different set of factors. For more information, please refer to Page 5 of this document.
We have noticed very low levels of correlation in daily returns between AiLA Products and other standard products within the Commodity and Equity space. Create a login here and access the correlations and data for analysis.
You can access it via ETF’s listed on Exchanges, Swaps offered by Investment Banks or Funds tracking our products. You may reach out to us at [email protected] to learn who are our partners and access those relevant tickers.