Rob Arnott: Unlock 20% Returns From Unloved Underdogs

During his recent interview on the Meb Faber Podcast, Rob Arnott discusses the inefficiencies in cap-weighted indexation, particularly around additions and deletions in indexes like the S&P 500 and Russell. Stocks added to indexes tend to be overpriced and underperform, while deleted stocks, often deeply discounted, tend to outperform by about 20% in the following year.

Arnott suggests a strategy of buying these “nixed” stocks, holding them for up to five years, which can result in an average annual outperformance of 5%. This approach captures value from small-cap, out-of-favor companies, offering a unique way to generate alpha from overlooked stocks.

Here’s an excerpt from the interview:

Arnott: And the aftermath was interesting. When we wrote that paper, we found that the additions, because they’ve been pushed up, by the time they’re added to the index, they’re very fully priced and they tend to underperform over the next year by a modest margin, 1 to 2%.

The deletions are less liquid stocks and the S&P indexation, for instance, owns 25% give or take of the market cap of every stock in the S&P 500. And so, when a stock gets dropped, 25% of its total market cap has to be sold in a single day.

So, those stocks on average bounce back by about 20% in the next year.

That seems likely to be kind of an interesting strategy. Now, the 20% is lumpy. It’s chunky. When S&P has had high turnover, like the aftermath of the global financial crisis during the run-up and crash of the dotcom bubble, turnover was high. And that’s also when a large value was added by the deletions after they were deleted.

So there’s turnover and it’s expensive turnover, there’s a 20% performance spread between additions and deletions. Well, gosh, that’s got to cost the indexes some money.

Multiply it by the turnover of the index, which tends to be about 4% per annum, and that works out to 4% turnover times 20 percent spread, you’d think 80 basis points. But, no, it’s not because the deletions are in fact tiny. We ask the question, what if you simply are patient about it?

What happens if when S&P changes the index, you stick a post-it note on your fridge saying this stock is being added, this stock is being dropped. And then a year later you take that post-it note and you do the trade.

If you do the trade 1 year late, you add over 20 basis points to S&P returns per annum compounded. Well, that’s pretty cool. So we began thinking about, is this unique to the S&P?

We looked at it on Russell. Very, very similar results. We looked at it on NASDAQ, NASDAQ 100, very similar results. We have done superficial examinations of, for example, EFA and other indexes, similar results.

That’s when we realized that it’s not that S&P or Russell are doing anything wrong or anything dumb, it’s just that by the very nature of cap-weighted indexing, you are deleting stocks that are deeply out of favor, trading at deep discounts.

And all they have to do is fare a little less badly than people expect, and the price can soar. So we decided, why don’t we create an index of deletions? We call it Nixed, and the paper points out that most of us at some point in our life have had the experience of being dumped. It’s not fun, but you can benefit and learn from it.

And likewise, when stocks get dumped, they can struggle and do very badly, or they can pull… get their act together and perform better than expected as a business and therefore perform much better than expected as a stock.

So we tested the idea of Nixed with S&P and Russell deletions. If you do that, what you find is that over the next 5 years, those deletions on average outperform by over 5% per annum for 5 years.

So, do you want to buy the deletions of this year and hold it for a year and have 100% turnover in thinly traded illiquid stocks, or do you want to buy a broader roster of stocks that have been deleted anytime in the last 5 years from, let’s say, S&P or Russell? Well, we want to be very respectful of their intellectual property. And so we decided to test our own index.

Let’s take the 500 largest market cap stocks in the United States and the 1,000 largest market cap stocks in the United States, and let’s look at how stocks that fall off of those indexes perform.

Well, gosh, they did exactly the same as S&P and Russell deletions. The value added was the same. So without encroaching on their intellectual property, we created an index, NIXT, that buys stocks that have fallen out of the 500 largest market cap or out of the 1,000 largest market cap stocks. And have done so by a margin big enough that they’re not likely to flip-flop back in on the very next rebalance.

And that very, very simple index beats the indexes that they’re dropped from by 5% per annum over the next 5 years. It also gets interesting in a number of other ways. Stocks that are dropped from the top 1,000, are they likely to be in the Russell 1,000? Well, of course not. They’re likely to be in the 2,000.

Stocks that are dropped from the top 500 by market cap. Are they likely to be in the S&P? No. They’re now small enough market cap that they’re likely to be out. And in fact, are reasonably likely to be outside of the top 1,000.

So interestingly, this sensible benchmark is Russell 2,000 Value. So the NIXT index is just a really interesting way of capturing small-cap value. And by using 5 years of deletions, you capture that full 28% outperformance. You wind up with an average of over 150 stocks, right now it’s 142 I believe. Stocks are removed after 5 years, or if they re-enter the index.

Now there’s a lot of flip-flops. When S&P and Russell drop a stock, oh, about a third or a fourth of the time it comes back into the index in subsequent years. When they add a stock it might turn out to be the next Tesla or the next Nvidia, or it might turn out to be some fluff stock that fails to live up to expectations and falls off the index a year or two later. ‘

There’s a lot of that kind of flip-flop turnover and that’s expensive. So, if you take the other side of those trades, you hold the stock for 5 years or until it re-enters the top 500 or the top 1,000.

In doing things in this fashion, you wind up with a broadly diversified portfolio. You wind up capturing Russell 2,000 Value returns plus that alpha, and it winds up being a great way to capture deep value, small-cap, out-of-favor, unloved companies. And everybody loves an underdog. So buying a collection of 150 underdogs, it’s a fun idea.

You can watch the entire interview here: