AI-Managed Portfolios

There is no universal strategy that will win in every scenario. Therefore, our performance is rooted in a vigilant, disciplined, tech-based response to market movements and evolving member needs - changing strategies, allocations, and even switching assets if needed to reduce the volatility of a portfolio. Based on our modeling, this continuous reduction of the “dips” enables AMPs to accrue incremental gains that add up, compounding over time. This is how AMPs aim to deliver better risk-adjusted returns in the long run than other automated products like ETFs, direct indexes, or investing with a robo-advisor.

AMPs take a systematic approach to risk management, using both technology and human expertise. This starts with you declaring what level of risk you are comfortable with to gain the expected returns. AMPs then model risk across several factors including individual and macroeconomic factors like sector and geographic exposure, interest rates, currency, and technical factors like momentum and reversal. They are trained using past performance, macroeconomic and company financial data and tested against alternative scenarios to ensure they are more resilient to black swan-like events. Finally, human risk experts with decades of experience and deep domain expertise provide heuristic guard rails as an additional layer of risk mitigation.

geographic exposure, interest rates, currency, and technical factors like momentum and reversal. They are trained using past performance, macroeconomic and company financial data and tested against alternative scenarios to ensure they are more resilient to black swan-like events. Finally, human risk experts with decades of experience and deep domain expertise provide heuristic guard rails as an additional layer of risk mitigation.

We believe that looking at returns as the only metric is overly simplistic and we aim to do this better. Nevertheless, the industry commonly talks about investing in this way and we keep getting asked how AMPs perform against established benchmarks, so here is a back test of our basic non-personalized AMP against an equivalent investment product (SPY) running in parallel at the same level of volatility.

BETTER

RETURNS¹

Annualized return over a 15 year period.

10.40%

AMP

8.92%

SPY

BETTER

SHARPE RATIO²

Higher sharpe ratios are generally considered “better”, offering excess returns relative to volatility.

0.615

AMP

0.510

SPY

BACKTEST USING $1m AS INITIAL INVESTMENT

1 Annualized Return (CAGR) over a 15-year period. Time period: from 1/1/2008 to 10/28/2022, pre-tax returns, net Arta management fees (the back-tested model assumes a fee of 10bp for a $1M portfolio. These fees are introductory pricing and are subject to change.) This basic AMP is tuned for improved performance over the long run based on accrued data; Arta will offer more tuning parameters as well as additional types of AMPs over time. SPY is an index that tracks the S&P 500.

2 Higher Sharpe ratio implies a higher return per unit of risk, where risk is measured as the annualized standard deviation of monthly portfolio returns. We calculate the Sharpe ratio as follows: To calculate excess returns, we subtract the risk-free rate derived from 13-week US treasury bills from the portfolio's annualized returns. To compute volatility, we annualize the sample standard deviation of monthly portfolio returns.

IMPORTANT DISCLOSURE INFORMATION: Back tested performance data is simulated by applying an AMP's machine learning model across ETFs spanning a broad set of asset classes, from January 1, 2008 to present. The machine learning model actively manages the portfolio to optimize for a target risk level based on market conditions at the end of every trading day. All returns are presented pre-tax and net of estimated management fees. The back-tested model assumes 10bp for a $1M portfolio. These fees are introductory pricing and are subject to change. Past performance does not guarantee future results; although our model may have outperformed a benchmark when applied historically, this does not mean that the same model will outperform the same benchmark in future market conditions. There are risks and limitations associated with the use of hypothetical and back tested performance, and such performance should not be used as the sole basis for investment decisions due to the fact that historical results cannot be duplicated.

2008 - 2011
Financial Crisis + Great Recession

During the largest market decline in the past 15 years, the AMP cushioned losses by allocating to defensive sectors such as consumer staples, utilities, and healthcare.

2012 - 2019
Bull market + low interest rates

The ability to diversify across a wide selection of ETFs enabled the AMP to deliver a higher return at the same risk level of the SPY.

2020 - 2021
Covid-19

While AMPs are designed to respond to market volatility and cushion losses, this version of our AMP experienced a greater loss vs. SPY during the month of Feb - Mar 2020. Early testing on our models in development show improved responsiveness with smaller drawdowns.

2022 Jan - Oct
Rising interest rates + high inflation

In the most recent, volatile period of monetary intervention and inflation, the AMP performed better than SPY, with a smaller drawdown year to-date.

Today, most people buy into a single generic strategy from a robo advisor or ETFs. These decade old solutions changed the way people invest, but they were based on the technology from the era when they were introduced and they tend to sell the same thing to everyone, regardless of personal preferences, goals, tax situation, and circumstances. A 30-year-old Bay Area tech worker shouldn’t have the same investment settings as a 50-year-old Houston based real-estate agent. That’s why every AMP is uniquely made for each Arta member.

You can tell your AMP what matters to you. AMPs take into account your preferences, including appetite for risk, sector concentration, geographic exposure, investment beliefs, or preferences for ethical impact. Over time, personalization will expand to factor in your overall financial position, where you live, and other aspects that should play into a comprehensive investment strategy. AMPs can enable you to deploy leverage or options to enhance returns while dampening risk. Whether you put an AMP on autopilot or fine-tune it to fit your needs, we put you in control.

Very few people have the time or the expertise to diligently manage their investments, day-in, day-out. AMPs are designed to automatically respond to market changes, maintain focus on investor goals, and help protect returns against excess volatility – all through the power of machine learning and technology. AMPs respond with minimal lag to market movements and life changes. By analyzing, reweighting, or even changing a portfolio dynamically, AMPs are able to maintain a stable risk level while staying focused on investor goals.

Arta’s proprietary ML models are at the core of AMPs. They systematically apply decades of financial research and regularly optimize across several asset classes, financial techniques, and tax implications. These ML models are trained and back tested on vast financial datasets including stock performance data, macro-economic indicators, and company financial metrics that don’t just respond to what has happened in the past but also simulate a broad range of possibilities.

The models are constantly evaluating their own level of confidence and learning in order to improve precision and performance over time. Unlike other financial managers, an AMP's actions are precisely tuned to the level of confidence it has in the decisions it makes. AMPs generate portfolios that solve for multiple changing priorities. So as financial markets and your life circumstances change, your AMPs automatically adjust to ensure your investments are always on-target with your long-term goals.

We believe that we win when our members do, so Arta does not charge for custody or trades. In addition, we apply these principles to our fee structures so that our incentives are aligned even more closely with our members:

◍

Transparent and fair: We’re committed to delivering pricing structures that are easily understood and contain no hidden fees. We will be clear and upfront with our pricing so you always know what you’re paying.

◍

Accessible for more people: active fund managers often charge the equivalent of 1% - 2% of AUM. Our extensive use of technology allows us to offer deeper personalization, higher responsiveness, and better risk-adjusted returns at price points that make sophisticated investing accessible to many more people.

◍

Choice of performance-based pricing models: We aim to put people back in control of their financial lives. Qualified members will be able to choose to pay based on performance (regulations restrict this to QC and above) so they can align our incentives even more tightly with theirs.