Asset Allocation

Leveraging machine learning to outperform traditional model portfolios and funds

Technology can remove the human bias in investing, whilst also offering investors exposure to the securities they own, but with reduced drawdown, minimized volatility and lower beta.

By Jeff Benjamin

It’s estimated that more than 80% of financial advisers are using some form of model portfolios to handle their clients’ investments. And in that overall category of outsourced investment management, the 10 largest models control about 75% of the market.

Beyond the case for safety being in numbers, any adviser looking to separate from the pack and show some value while still outsourcing might want to look beyond the beaten path of model portfolios.

With that in mind, we’re taking a look at UX Wealth Partners, a relative upstart turnkey asset management platform that has grown to $1 billion since launching in 2019.

UX has many of the same bells and whistles as any other tamp, including a broad range of investment options and models to choose from.

But the stuff chief executive and co-founder Kyle Wiggs wants to talk about is the strategies relying heavily on quantitative analysis and artificial intelligence.

Four of the asset managers on the platform are combining to produce chart-topping relative performance across time periods ranging from one month to three years.

Three of those asset managers — AI Funds, Qraft Technologies, and StockSnips – are managing portfolios with variations of artificial intelligence and machine learning.

The fourth, Thor Financial Technologies, of which Wiggs is a minority owner, employs a digital signal processing technology that is analogous to the electrical engineering science that neutralizes sound waves.

“They’re doing it from an ETF perspective to determine what is material and what is meaningful in the financial markets,” Wiggs said.

Patterns and probabilities

Without getting too deep into the weeds of the science, it is worth looking at some of the results.

For example, one model portfolio, AIQQAI, just responds to patterns and probabilities for each coming week.

If the outlook is bullish, the model will allocate 100% to the Invesco QQQ ETF tracking the Nasdaq 100 Index.

If the outlook is bearish, the model will move to cash, and if the outlook is neutral, the model will hold half cash and half QQQ.

That strategy is down 1.6% this year through June 24, which compares to a decline of more than 28% by QQQ and a decline of nearly 18% by the S&P 500 Index over the same period.

The UX platform offers the same model strategy for the Dow Jones Industrial Average and S&P, and the relative performance is equally impressive.

“Our overarching goal is to tap into technology driven solutions that eliminate the human component and give investors exposure to the securities they already want to own, but with reduced drawdown, minimized volatility and lower beta,” Wiggs said.

Another example of how the technology is applying risk-on and risk-off exposure based on probabilities is illustrated in the Next Gen ETF model, which is designed to beat the ARK Innovation ETF (ARRK).

The Ark fund’s three-year annualized return through June 24 was 3.6% with a standard deviation of 43.6%.

Meanwhile, the Next Gen ETF model portfolio on the UX platform had a three-year annualized return of 18.9% and a much lower standard deviation of 23.7% because the beta to the Ark fund was 0.48.

Wiggs said his inspiration for developing the UX platform was realizing that this corner of the wealth management industry wasn’t utilizing technology to its fullest potential.

‘Everything was the same’

“I spent 15 years in the traditional asset management space, and everything was the same,” he said. “The same managers and the same results.”

By embracing technology, Wiggs said he is removing the potential for “human bias.”

“We’re talking about using technology instead of humans to make investments in real time, because traditional buy and hold strategies are not going to make it going forward,” he said.

“What we’re doing is not buying alternatives, we’re buying all the things already in the portfolio, but we’re taking the approach that technology can detect opportunities better and act quicker.”