arty’s Investment Approach

arty’s objective is to maximise your return for your chosen level of risk. We measure risk using volatility, or the amount your annual portfolio return may deviate from the average return. To achieve this, arty selects a combination of ETFs (for Tailored customers arty can also include your existing stocks) and weights each constituent in the portfolio using our unique Artificial Intelligence optimisation technique. This technique is proven to deliver sustained capital growth over time and outperforms the equivalent, equally-weighted portfolio.

To describe this process, we will cover three topics: Risk, Return and Yield.


arty assesses every customer’s attitude to risk by asking them to complete a risk tolerance questionnaire (you can revisit the questionnaire at any time). This yields a risk tolerance value of between 0 – 100. Based on this value arty sets a target volatility for the customer’s portfolio. Target volatilities range between 4 – 25%.

Suppose the average return is 10% and the target volatility is 10%. This means, in practice, that there is a 66% chance (1 standard deviation) that the expected return of the portfolio will be within 10% of the average return. In other words, the return is likely to be somewhere between 0% and 20%.

Editor’s note – volatility is not necessarily bad in that it does not per. se. reduce the expected return of a portfolio. However, you should be careful with volatility if you need to sell off your investments at a particular time. A more volatile portfolio has a higher chance of a significant deviation from the mean on any particular day/month. Also, it has been shown that human investors are more likely to panic when their investments fall and sell them off at lower prices. A less volatile portfolio reduces the temptation to do this. On the other hand, it is important to note that targeting a higher return usually involves accepting a higher level of volatility.

arty has two ways of targeting volatility. Firstly, arty can adjust the monthly weightings within the portfolio. For example, to reduce the overall volatility, more volatile assets will receive a lower weighting, and less volatile assets will receive a higher weighting—more on this below. Secondly, arty will choose a mix of ETFs based on their overall correlation. The lower the total pairwise correlation between the ETFs, the lower the overall volatility will be because the portfolio constituents will not all be moving together. This is another reason why arty’s portfolios may include an allocation to the ART fund, which has a very low correlation with traditional assets. 

Note – the ART Fund is an actively managed systematic investment fund. It invests in futures markets using an automated, AI-driven investment algorithm. To find out more, please visit


Once arty has the target volatility, the next objective is to maximise return. This is again down to ETF selection and weighting.

arty will select ETFs for each person based on their return potential, as measured by the backtesting performance of an optimised portfolio made up of those ETFs. As a general rule, equities have a higher return potential than commodities and bonds.

Backtesting is used to find the best ETFs (from a pre-selected universe) and the ideal weights for each constituent. This is where Artificial Intelligence is particularly effective. Each weight is selected based on several testing parameters. arty will run many simulations, learning and adjusting the parameters and weights until the perfect weights are found. This is done for many possible combinations of ETFs (while still maintaining the target volatility).

The objective of the backtesting is not to find the portfolio that delivers the highest return over the previous period. This is known as overfitting and is not an effective way of ensuring future returns. Instead, the objective is to design tests, that choose weights that deliver the highest monthly returns over the backtesting period. Effectively, we break the backtesting period into months. At the beginning of every month we try to design tests that will yield weights that maximise the upcoming monthly return based on the available information. When repeated over many months, this produces robust tests that can then be deployed in live portfolios.

This process is repeated every month for every customer using the latest available data. This way arty is always learning, improving and reacting to the latest information from the markets.

As an arty customer, your portfolio will be rebalanced monthly to ensure you continue to achieve the maximum return for your target level of volatility. arty will then show you how to adjust your current trades / stocks to reflect the new weighting.


It is important to mention that while some investors target yield (e.g. dividends or bonds) and some investors target capital gains (e.g. single stocks), arty targets both. In practice, this means that arty selects a mix of ETFs with different characteristics. There are high yield corporate bonds, dividend-paying stocks and growth sectors such as commodities and thematic equities, e.g. cybersecurity or robotics. Also included is the ART fund, which targets absolute returns. All ETFs are accumulating, which means that any dividends or coupons earned are automatically reinvested, effectively converting yield into capital appreciation.

By optimising a portfolio of accumulating ETFs, arty can incorporate yield into the analysis, thereby targeting the highest overall level of return.


Unfortunately, there are no guarantees in life and especially so with investing. However, by using the latest techniques, technology and artificial intelligence, arty can offer investors the chance to control their risk and achieve a return that would be unlikely with most other forms of investment.

Keep reading

Access for free ETF Quant portfolios

arty ETF portfolios are based on AI