VettaFi recently sat down with the team behind the Hull Tactical US ETF (HTUS B-), which employs an innovative approach to tactical investing. The full standardized performance, through September 30, 2026, is shown in the graphic below. VettaFi discussed the fund’s quant-heavy approach and its role in portfolios with Hull Tactical Financial Engineer Aishvi Shah.Nick Peters-Golden: What does this fund aim to accomplish? What is its investment thesis and how does that differ from your standard S&P 500 index fund?
Aishvi Shah: We believe investors should stay invested for the long term, but their equity exposure can adapt as market conditions change. HTUS uses quantitative models that combine dozens of market signals to adjust its U.S. large-cap equity exposure. Unlike a traditional S&P 500 index fund, which generally stays fully invested, HTUS can range from 0% to 200% long. Our goal is to participate in market gains while reducing exposure when our signals weaken.
On average, we are 100% long the market, because we fully believe that the market tends to go up over time. We think that stays true. So that’s how we got started. And then, based on a quantitative model that uses 40+ indicators to forecast near-term S&P 500 performance, the fund either dials up or down its overall market exposure, both at the beginning and end of each trading day.
We are very much an actively managed fund. And even in the space of actively managed funds, I would argue that we’re active-plus. Generally, actively managed funds tend to have a heuristic that they trade to every day or every quarter or every month.
Our fund is a little different because we are dynamically adjusting our active strategy on a day-to-day basis, based on what’s happened the previous day or what’s happened even on the day of.
Peters-Golden: How is your model built, and what is it designed to predict?Shah: I think for me to explain how our model’s built, I have to start with micro alphas, which is a paper that both our PMs and our founders worked on. It is definitely available on our website, so people can find that.
The thesis is that you take signals, which can be indicators that are just well known. Everyone would know about these. They’re different sentiment indicators, valuation indicators, and maybe their signal’s a little too weak to be statistically significant or actively tradable. But we found that if you combine enough little signals, you can get meaningful predictive power.
We have about 40 to 50 different features in our models that have micro alphas in them, which we add together then to create our real alpha. In terms of what we predict, we are always looking to predict the next day’s return, and that’s because we are always trading every day.
Peters-Golden: Are you aggregating these indicators, or are there some micro alpha indicators that are stronger than others?
Shah: I would say yes to both of those questions. So we are aggregating them, and that’s a little bit where our math background comes into play — that would give us a little bit more advantage over the retail investor that might be trying to play around with these. But in that aggregation, of course, some get a stronger weight than others, based on if that signal has a micro alpha that’s a little stronger than the other ones.
Peters-Golden: As a follow up, how does your research process turn individual signals into a single trading decision? How often are you tuning your weights for these micro alpha signals?
Shah: So, we are tuning that weight every single day, technically refitting our model. That being said, that particular weight is not changing as quickly, because you’re really only adding in an extra day.
In terms of even just a single feature, one micro alpha, the construction of that is a little more nuanced too, because the parameters used to create that one feature could change and that could change based on different market regimes that we’re in.
So if we, as portfolio managers, are looking at it and we say, “Oh, we have a different market regime,” we may want to reset that parameter, take a longer moving average, take a shorter moving average — that number can change.
Peters-Golden: With AI making it easier to analyze public data, how easily can this fund be replicated?
Shah: So, we’ve actually tested this out. About a year and a half ago, when AI really started booming and there was a lot of generative AI software for sale, we bought a trial into one of them. We said, “Here’s all our data. This is what we’re trying to predict. Do what you can with it.”
We had a very tough time with that. One thing generative AI struggles with a lot is overfitting, because it’s very easy. If we give you an in-sample training period and a test period, you can mangle that training period as much as you want to get your test to look great.
We even more recently went into a Kaggle competition, where we gave them a bunch of data and said, "Look, we want you to beat our model that we trade on. So everything that we have, we’re giving you. Please try to beat it.”
We were hit with a very similar issue: You can beat it in the six months, but quite often that signal doesn’t hold up.
That’s something when people ask, “Where is the whole secret sauce? What are you worth?” And it’s our track record, to show that we’ve been doing this not just six months, not just one year.
Peters-Golden: How has it evolved over time then? We were talking about track records. There’s also a track record of how efficiently you’ve updated your model, not just in the numbers of how it’s actually delivered, but in your ability to adapt.
Shah: One thing that makes it easy for us is that we never stop researching. In terms of how we’ve been able to keep our track record, we’re constantly evolving not only the model that we built, but [we are also] adding different models.
When we first started out, we used to be more of a long-term model. We used to forecast one month or six months out. And as the market has shifted, you can see that even in terms of just more trading volume, lower expiries, options coming into the market, the market reacts a lot faster to information.
We started adding in more one-day models. [However,] in the past year, we’ve looked at bringing back some of our longer horizons. Maybe not as long as six months, but we’re looking more into five-day, one-month models.
*Peters-Golden: Could you elaborate a little bit on how you go from having the perfect prediction and then converting that into portfolio decisions, at least on what you’re holding, versus how you account for risk? *
Shah: What our prediction basically tells us is how confident we are that the market’s going up or down tomorrow. We’ve found that you can beat the market. We think the market’s going up on average. If you just take a 2X levered position, when you think the market’s going up, you could be right.
The biggest problem is risk. We have our volatility constraint. If people wanted to just be in a 2X levered fund and make money, they could do that themselves. That’s not what we’re here for. We’re here to offer extra alpha, but [with the goal of] still staying below the volatility of the S&P.
Peters-Golden: Recent bond market selling has reminded investors that bonds carry risk too. It’s been a long time, but we’re back in this regime now. How does actively adjusting equity exposure help advisors manage portfolio risk and what are the limitations of that?
Shah: You want to adjust your equity exposure to your risk tolerance. Now, what I have to say in general in terms of advisors holding different portfolios, say a 60/40 or a 70/30, historically we’ve seen in high volatility times, cross-asset correlation has gone up, and which is why we’ve seen that, historically, a 60/40 portfolio does underperform.
We factor in rates, valuations, [similar] things into our models, and so that’s one of the factors that goes into us feeding the S&P. So, I would say [that advisors] should factor that in. If they’re holding HTUS in their portfolio, it would let them be a little less risk averse in these times.
*Peters-Golden: When enthusiasm around AI stocks is strong, how do you stay disciplined if it calls for less equity exposure? And how should advisors explain that positioning to clients? *
Shah: I think this has to do with what I was mentioning in terms of an emotional reaction to things you see in the market. We’ve seen it happen this year.
But it’s important to note: The market is not often just driven by one thing. This is where I think the model does a pretty good job of aggregating everything, instead of just making decisions based on the top headline in the news.
Overriding the signal is just not something we do here. So when an advisor calls with stress, things like that, we like to say, “Don’t look at your returns on a day-to-day basis. Try to get it more aggregate.” But like I said, it’s a long-term holding. We just follow what the model says and stay disciplined.
Peters-Golden: Finally, how does this fit into a portfolio?
Shah: All advisors have a large-cap position somewhere in the market. And we know that we’re a very new product. It can be intimidating. We’re happy to talk about the math as much as people want, [but] it’s new to a lot of people.
For more news, information, and analysis, visit the Equity ETF Content Hub.DisclaimerCarefully consider the Fund’s investment objectives, risk factors, charges, and expenses before investing. This and additional information can be found in the Fund’s prospectus, which may be obtained by visiting www.hulltacticalfunds.com or by calling toll-free 1-844-484-2484. Read the prospectus carefully before investing.
HTAA, LLC serves as the investment advisor. The Fund is distributed by Northern Lights Distributors, LLC (225 Pictoria Drive, Suite 450, Cincinnati, OH 45246), which is not affiliated with HTAA, LLC.
About the Hull Tactical US ETF (HTUS) Investment Strategy
HTUS is an actively managed exchange traded fund (ETF) driven by various proprietary analytical investment models that examine current and historical market data to attempt to predict the performance of the S&P 500® Index (the “S&P 500®”), a widely recognized benchmark of U.S. stock market performance that is composed primarily of large-capitalization U.S. issuers. The models deliver investment signals that the Adviser uses to make investment decisions for the Core Equity Replacement Income Generation Hedge Fund Style Research in ETF Wrapper May serve as replacement to portion of client S&P 500 exposure Integrated option strategies designed to generate income Active research with foundations rooted in academia Fund. The investment models used are to anticipate forward market movements and position the Fund to take advantage of these movements. Currently, signals are combined into an ‘ensemble’ array that spans statistical, behavior-sentimental, technical, fundamental, and economic data sources. This combined signal is generated each trading day towards the close of the market and dictates whether the Fund is long/short and the magnitude of position sizing. The Adviser routinely evaluates the performance and impact of each model on the Fund with the goal of realizing a risk/return profile that is superior to that of a buy and hold strategy.
The use of derivative instruments involves risks different from, or possibly greater than, the risks associated with investing directly in securities and other traditional investments. These risks include (i) the risk that the counterparty to a derivative transaction may not fulfill its contractual obligations; (ii) risk of mispricing or improper valuation; and (iii) the risk that changes in the value of the derivative may not correlate perfectly with the underlying asset, rate, or index. Derivative prices are highly volatile and may fluctuate substantially during a short period of time. The use of leverage by the Fund, such as borrowing money to purchase securities or the use of options, will cause the Fund to incur additional expenses and magnify the Fund’s gains or losses. The Fund’s investment in fixed income securities is subject to credit risk (the debtor may default) and prepayment risk (an obligation paid early) which could cause its share price and total return to be reduced. Typically, as interest rates rise the value of bond prices will decline and the fund could lose value.
The Fund is an actively managed ETF and, thus, does not seek to replicate the performance of a specified passive index of securities. The Fund may take short positions. The loss on a short sale is theoretically unlimited. Short sales involve leverage because the Fund borrows securities and then sells them, effectively leveraging its assets. The use of leverage may magnify gains or losses for the Fund. There is no guarantee that any investment strategy will produce positive results. There is no guarantee that distributions will be made.
The thoughts and opinions expressed in the article are solely those of the author. The discussion of individual companies should not be considered a recommendation of such companies by the Fund’s investment adviser. The discussion is designed to provide a reader with an understanding of how the Fund’s investment adviser manages the Fund’s portfolio.
The Advisor has implemented policies and procedures to mitigate potential risks associated with the use of machine learning or artificial intelligence, including development, use of, supervision, and testing of models and algorithms.
