By From the bustling days of open outcry—where traders and brokers shouted buy and sell orders on the trading floor—to today’s seamless “click-and-transact” mobile platforms, equity trading and investment have evolved significantly in India and across the world.
Now, like many industries, equity investing is being reshaped by artificial intelligence—one model at a time.
With the rise of AI-driven investment platforms, India’s market regulator, the Securities and Exchange Board of India (SEBI), is closely monitoring this rapidly expanding space.
One such platform, InvestorAi, founded in 2018, leverages AI to provide stock recommendations, aiming to carve out a strong presence in investor portfolios—particularly among Gen Z and millennials.
After securing ₹80 crore ($9.6 million) in Series A funding from prominent market investor Ashish Kacholia last year, the platform is setting ambitious goals.
It plans to expand into mutual funds, asset classes like commodities and gold, and even international portfolio recommendations.
In an interview with Invezz, co-founder and CEO Bruce Keith likens AI-powered investing to “giving investors an Iron Man costume.”
He explains, “You’re still in control, but you have these superpowers alongside you.”
Keith also offers insights into how much of a portfolio should be allocated to AI-driven investments, why transparency in win and loss rates is crucial for investor trust, and why he believes AI-powered stock advisory will command a third of the market in the coming years.
Edited excerpts:
The goal of InvestorAi
Invezz: How did InvestorAi come about and what the key inflection points have been since the Indian stock market has grown significantly over the last decade…
Our goal from the outset was to democratize wealth management.
In my previous business, we worked in the hedge fund sector, providing risk management solutions.
We saw how hedge funds leveraged AI and ML in their strategies. And, there is always an imbalance, isn’t there?
The kind of institutional investor or the hedge fund investor is always going to have more power than the little guy.
So, one of our main objectives was to level the playing field.
Out of my two co-founders, Sarthak comes from an AI background and approached the challenge with an AI-first mindset.
Our third co-founder was leading Barclays Wealth globally at the time.
Therefore, we had this combination of skills coming together.
When we look at the wealth management market, there are plenty of services available—but not everyone wants them.
Some investors prefer a hands-off approach.
On the other end there are digital investors, which, in the old days, we used to call them self-directed or DIY investors who managed their portfolios.
I am a DIY investor, and I do it all on my phone. For millennial and Gen Z investors – this is their go-to as well.
Besides, in India, platforms like Zerodha, Groww, and Upstox have done an excellent job of providing market access.
For me, democratization has three key aspects.
The first is education—media outlets, platforms like CNBC-TV18, and online communities have played a crucial role in enhancing financial literacy.
While influencers and social media often get a hard time, many provide valuable insights.
Platforms like Reddit also get a hard time, but communities that add to the education component are good.
The second aspect is access—newer brokerage firms have significantly streamlined market entry.
India, in particular, is ahead of the curve, thanks to digital infrastructure like eKYC, Aadhaar, and other advancements.
Opening an account in India is incredibly fast—at one point, the longest it took was around seven minutes, maybe 11.
In contrast, in countries like the UK, the process can take several days.
So the whole India stack, the whole DPI piece, has made a massive difference in kind of helping people easily get into markets.
‘AI like an Iron Man costume’
So you’ve got some education, you’ve got people in the markets, it’s now about how you kind of give them the kind of right experience off the back of that.
And, you can think of AI as this kind of amorphous, big, bad thing out there.
Or, you can think of Robert Downey Jr. and an Iron Man costume.
What we’re trying to do is give people an Iron Man costume.
So you’re still in control, but you’ve got all these kinds of superpowers along with you.
We’ve been building our AI since 2018 and running portfolios since the pandemic.
Our goal is to provide investors with an AI-driven edge.
By subscribing to our baskets, investors can benefit from AI-driven insights while maintaining autonomy.
This, to me, is true wealth management democratization.
You buy a basket that rebalances every month or you buy one that is intraday, it doesn’t matter.
I don’t think anyone should have all their money invested in AI.
For me, it’s a systematic investing part of your portfolio.
And, our track record over the last three and a half, four years has been pretty strong.
How does InvestorAi work?
We use AI to generate model portfolios or equity baskets.
The user sees the recommendations and presses the button.
The user presses the button, and then the transactions happen all at once.
It transacts in one go. So, they don’t have to go and type out 20 different orders, it’s a single-click execution.
However, users still can exclude certain stocks, modify weightage, or choose not to invest in specific securities—giving them more control.
Users can’t add to the basket, but you can remove stocks or adjust the weightage.
The same applies to our intraday and MTF baskets.
The core idea is that users remain in control.
This is partly for regulatory reasons but also aligns with what we sought to do at the outset.
InvestorAi has outperformed the market by 2x
Invezz: Can you quantify success rates? I know investing outcomes can be difficult to measure.
So, you can measure success.
In simple AI parlance, there are two metrics- the win rate, which is did they make money? So, that’s, kind of zero or one.
You can then measure the beat rate, which is, did they do better than the market? Again, zero or one.
So, those are simply our two kinds of simple metrics.
And, I genuinely believe that anyone who’s in the investing market using AI should be publishing win rates and beat rates.
Over the past three and a half years, we’ve generally outperformed the market by about 2x.
Of course, this hasn’t been a straight line—there have been ups and downs, as with any market cycle.
Intraday’s interesting.
So, part of the reason we got into intraday isn’t because we can make a load of money on it.
It’s because according to SEBI, only 10% or 9% of people made money on intraday.
That’s horrible, right? Why would you do it if it’s a kind of 1 in 10 or 10% chance you’re going to make money?
So, last year our intraday was running at about 55%.
We’ve just put some new things through at the start of this year.
Since we put them in, we’re now running at about a 75% win rate.
InvestorAi and intraday trading
Invezz: So, is intraday where this model blooms?
No, honestly. What intraday showed us—so, here’s the thing with AI.
AI is great when you’ve got lots of data.
Humans are great when we don’t have much data because we apply context and experience.
AI is your idiot savant—you give it one task, and it is deep and narrow, doing its thing.
AI beats humans when a lot of data makes a difference.
Humans, on the other hand, currently outperform AI when there’s less data and connections need to be made that aren’t obvious.
Whether it’s intraday, a two-week rebalance, or a one- to two-month rebalance, two months is the longest we have for any portfolio.
The higher-return portfolios tend to have a two-week rebalance period, but some investors prefer less frequent adjustments.
If you don’t rebalance, you might miss out on the next opportunity, so it becomes a matter of investor choice.
When a notification comes through, do you want to act on it, or would you rather rebalance it only once every two months?
AI is everywhere, and there’s no reason why it shouldn’t be part of investments.
It should help people, but it works best for those who want to engage with it. If you prefer a passive approach, a two-month rebalance cycle is better.
If you’re comfortable with a faster pace and willing to act when notified, you can benefit from crunching additional data.
The current user base
Invezz: What’s your user base like? Are you specifically targeting new or young investors? Could you quantify that?
Ironically, our current user base consists mostly of established investors.
We operate entirely through brokers, not B2C.
This was a deliberate decision for two reasons. First, acquiring customers through brokers keeps costs low since they already have engaged clients.
So, it’s going to pre-screen for us and we’ll do a revenue share with the broker.
Second, the percentage of active DMAT accounts in India is about 20%.
That’s terrible.
In the UK, it’s around 50%, and it’s in the high 40s in the US.
While opening an account has become easier, engagement has declined.
Many traditional brokers want to re-engage clients, and this type of AI-driven investing helps them do that.
Our client base includes HDFC, Yes Securities, IIFL, Prabhudas, and Geojit.
We’ve also recently launched with 5Paisa and are about to go live with HDFC Sky and a neo-broker platform.
We’ve started creating offerings that attract Gen Z investors, but ultimately, being a digital investor isn’t about age.
It’s about wanting to manage investments yourself. That’s where the biggest benefits will come, and I’m excited about this next stage of growth.
Triple-layered approach for real-time market changes
Invezz: Currently, Indian markets have corrected, so how do you factor in the changes?
So, it has been a couple of months, but actually, the market’s not down that much.
I was working in private equity during the dot-com boom and also during 2008.
I’ve seen some horrible market moves.
So, what’s happening now isn’t a market fall.
It’s not even a correction; it’s just a small pause or dip.
I think, fundamentally, India, as a market, is the place to be for the next, comfortably, 20 years to 30 years, probably similar to how the S&P performed from the 1990s onward.
There will be dips, but overall, it’s a bullish market.
And that’s where the nifty is for me today.
If I look at my investments, and, a good chunk of my wealth in India, I see it just as a good, long-term bet.
In terms of portfolio adjustments, yes, we continually tweak and refine models based on market conditions.
It’s not a static process but one that updates in real-time.
But what we’ve ended up doing, is we’ve got, core models and we’ve got contender models.
So, if one of them is underperforming, we’ve got one that’s already running, that can switch in.
But we also have, sitting over the top some meta-models which are interesting.
Some act as market barometers, helping determine trading decisions.
For instance, in intraday trading, a top-layer model may decide whether to only take long or short positions based on market conditions.
These meta-models respond to factors like rising volatility or an increase in the VIX, prompting a shift to more defensive strategies.
Essentially, it’s AI managing AI, enhancing overall performance.
On SEBI’s monitoring of the AI-driven trading
Invezz: What are the risks of investing through AI-driven models? SEBI is also keeping a close watch on how this space is evolving.
Fundamentally, I think SEBI is doing all the right things. I understand what they’re doing in terms of algorithmic trading even though it’s not the world that we’re in.
I think there’ll be a bit of rework as they get more feedback from participants.
If I think about the AI trading world that we’re in, where people are using AI alongside it, you know, rightly, you’ve got to be a research analyst and you’ve got to have some human oversight on what comes out before it goes to the customer.
And you’ve got to be able to explain to the customer why you think this is a good stock.
Now, it’s difficult to explain the AI down.
So, you’ve got two forms of AI.
Historically, you had symbolic AI, which was all about logic and reasoning.
So, you set lots of rules and the machine runs on the rules.
Our stuff’s all built on neural nets, and neural nets are trying to mimic the human brain.
So, what you’re doing there is you’re telling it what the outcome is and you’re allowing it to make the connections itself to find its connections.
So, you can’t tell precisely how it’s made the connections in terms of a company coming out.
But what you can do is you can screen the companies when they come out to ensure that it’s a good stock according to you, and you’ve got a good reason behind it.
SEBI insists that a human piece sits over the top.
And actually, that works. So, we run our stuff. We put that over the top.
Before it goes to the customer, it’s got that.
The industry needs to publish win/ loss rates
I think the one thing I would like to see SEBI do, and, not SEBI, the industry, I think, must do, is could it be more transparent.
If you’re more transparent about your win rates or your loss rates, or your beat rates, or whatever the opposite of beat rate is, and everyone reports it on the same basis, then it makes it much easier for people to start to trust the model and the process and the system.
And, so much of the talk around AI has to revolve around trust.
We’re trying to take positions on some of this stuff and starting to publish data consistently.
As more people start to do what we do, I think SEBI will start to insist on certain criteria that are measured and that you’re clear on how you measure them.
I think that’s just a given in terms of where we’ll end up. And I think that’ll help a lot.
You know, in the same way, that, when you look across mutual funds, you’ve got to be able to compare the Total Expense Ratios (TER) and different mutual funds.
If everyone does it differently, then it becomes really difficult for the consumer, and for the investor.
Expanding into mutual funds, other asset classes?
Invezz: Right now, you offer equity baskets and portfolios. Do you have plans to expand into mutual funds?
Absolutely. Our vision is to democratize wealth management.
Right now, we primarily serve active investors, but that’s only a segment of the overall market.
The short answer is yes—we have this in our business plan, though it’s still a few years away.
The key will be choosing the right partners.
Just as we have partnered with brokers to reach DIY investors, we will need trusted partners with strong distribution networks to enter the mutual fund space.
Our strength lies in AI and intelligence, not mass distribution.
Finding the right collaborators will be crucial.
Another avenue for expansion is into other asset classes.
Invezz: Could you elaborate? Also, does crypto fit into your broader strategy?
We’ve done some crypto stuff before we stopped because it was just too difficult from a regulatory perspective.
But most of my team members have some crypto investments.
We’ve all kept some and kind of done okay off the back of that.
So, it wouldn’t surprise me if we were back in there.
But from an Indian context, everyone loves gold.
So across the commodity piece, in particular precious metals, in India it is better understood than in many other places.
And they don’t all move in sync. Therefore how you play in and out of that market becomes quite interesting.
We’ve got some international equity stuff running just now.
Not available in India, but we’re working with a fund to put something together outside of India on that.
And I think that I can see the kind of international piece coming into play.
The one place I don’t see AI having a massive advantage is when it comes to private assets.
Because there’s not enough data.
Like physical real estate, some of the amount of data that’s there and just how you and the right quality of data that’s there that allows you to do that.
I just don’t think it sits very well with what we do. So it has to be something where there’s a good degree of trading going on.
On building international portfolios
Invezz: When you mentioned international equity, I wasn’t quite clear. Are you referring to domestic investors wanting to invest in Indian international equities, or international partnerships you’re entering?
A bit of both.
We’re currently working on a partnership with an international fund to create an international portfolio.
That partnership brings a high degree of credibility, which is valuable. Beyond credibility, they want us to handle the international segment.
Once that’s in place, we can start creating portfolios that an LRS investor could use.
I believe it’s only a matter of time before those become available.
However, the more interesting question is how to create compelling investment products, such as ETFs, for those wanting to invest in India.
It’s quite difficult for someone outside India to invest directly in the market.
For example, my kids are in the UK, and their options are limited to ETFs tracking the Nifty 50 or Nifty 500.
There’s potential to expand these options.
But there’s a piece there around, again, if Gift City becomes the point of access, how do you create the kind of engaging products that kind of sit alongside that?
AI-driven investing could make up a third of the overall market
Invezz: Looking ahead to five to ten years, what mix do you foresee between AI-driven and traditional financial advisory? How much of the traditional space will AI replace?
Let me give you a number—though it may not be exact, it helps as a reference.
Let’s look at the US as a proxy and look at active funds, passive funds, and quant funds.
The US in 1989 is roughly where India was in 2023. The market was 89% active funds, 11% passive, and had no quant funds.
India in 2023 was 82% active, 17% passive, and 1% quant. I know quant and AI are different but they are kind of similar.
By 2023, the US had shifted significantly—active funds dropped to 26%, passive grew to 39%, and quant funds surged to 35%.
If I just take that as a proxy, if India follows the US trajectory, AI-driven investing could see at least 35x growth in mutual funds alone.
And, the market’s going to be growing as well, but in terms of market share, if you assume new market growth, it’s going to be a third of the overall piece.
In my portfolio, for instance, about 25-30% is allocated to systematic investing.
I also hold long-term ETFs and some yield-focused stocks and commodities.
I expect AI-driven investments to eventually represent around a quarter to a third of the retail market.
The post Interview: AI for stocks is like giving investors an Iron Man suit, says Bruce Keith of InvestorAi appeared first on Invezz
