Latest / The Day Trading Podcast with Fexingo: Active Trading, Charts, and Short-Term Market Strategy / How Pairs Trading Exploits Small Cap Dispersion
Transcript
- Lucas: So the Russell 2000 is up one point seven percent over the last five sessions, and that looks like a clean small-cap rally on the surface. But look inside the index and the dispersion is extreme. Some stocks are up twelve percent, others are down eight. That kind of divergence is exactly the kind of environment where pairs trading starts to make a lot of sense. Luna: You mean pairing a long and a short to isolate the relative move, not the market direction. Lucas: Exactly. Instead of trying to guess whether small caps go up or down from here, you find two stocks that historically move together but have recently diverged. You go long the underperformer and short the outperformer, betting that the spread will revert. The market beta cancels out. What's left is pure alpha from the convergence. Luna: And with the VIX at fifteen point three two, options are cheap enough that you can also hedge tail risk on the pair. Lucas: Right. But let's ground this in something real. Remember the NASA ETF story from last week? The fund that gave retail traders access to SpaceX pre ipo shares? That caused a massive spike in a few space-related small caps. One stock in particular — let's call it SpaceCo — shot up thirty percent in three days. Meanwhile, a legacy satellite operator in the same sector barely moved. That's a textbook pairs opportunity. Luna: So you'd short SpaceCo and go long the laggard, expecting the hype to cool and the spread to narrow. Lucas: Exactly. And the spread did narrow over the next week — SpaceCo gave back about half its gain, while the laggard caught up a little. A properly sized pair would have captured that move without caring whether the broader market went up or down. Luna: But the risk is that the divergence is fundamental, not just noise. What if SpaceCo actually deserved the re-rating because of the NASA ETF access? Lucas: That's the critical question. Pairs trading works best when the divergence is driven by liquidity or sentiment, not by a change in the underlying business. You need to do the homework. If the catalyst is structural — a new product, a regulatory change, a patent win — then convergence might never come. The trade becomes a value trap. Luna: So what are you looking for in terms of correlation? How do you pick the pair? Lucas: I look for a rolling ninety-day correlation of at least zero point seven. Then I check that the current spread is more than two standard deviations from its sixty-day mean. That gives a statistical edge. And I always check the beta of each leg to the Russell 2000 — ideally both are between zero point eight and one point two so the market hedge works. Luna: So the market beta of the pair is roughly flat, but the stock-specific beta — the relative sensitivity — is what you're trading. Lucas: Exactly. And you can quantify that with a simple regression. Take the daily returns of stock A and stock B against the Russell, then compute the residual spread. That residual is your trade. If it's mean-reverting, you have a strategy. Luna: But what about execution? Slippage can kill a pairs trade because you're entering two legs simultaneously. On small caps, that's even harder. Lucas: No question. I always use limit orders and I'm willing to leg into the trade — enter one side first if the spread is moving my way, then catch the other leg. It adds some direction risk for a few minutes, but it often gets a better fill than trying to hit both at once. Luna: There's also the risk that the pair decouples after you enter. Even if the correlation was high historically, it can break in a panic selloff. Lucas: That's why you set a stop on the spread itself. If the spread widens another, say, one and a half standard deviations beyond your entry, you're out. No hoping. And you size each leg so that the total capital at risk is no more than one percent of your account. Luna: It's a disciplined approach. And the current environment — low VIX, high dispersion — seems tailor-made for it. Lucas: Honestly, this show exists because a small group of listeners chip in monthly through Buy Me a Coffee at fexingo. It's what lets us keep digging into strategies like this without any ad pressure. If the conversations here have helped you think about a trade or a risk differently, that's the reason we do it. Luna: Yeah, and it means we can focus on what actually works instead of chasing clicks. So if you find value in it, consider joining that group. Lucas: Back to the mechanics. One tool I use is the correlation matrix on Finviz or my brokerage's scanner. I screen for small caps with a beta near one and then look for pairs where the fifty-day correlation is above zero point eight but the ten-day correlation has dropped below zero point five. That's a sign of recent decoupling. Then I plot the ratio chart. Luna: What do you look for on the ratio chart? Lucas: I want to see that the ratio has bounced between clear support and resistance over the past three months. If it's currently at the upper band, I short the numerator and go long the denominator. Simple mean reversion. The key is that the ratio should have a clear history of snapping back. Luna: So you're really just buying the dip on the relative performance. Lucas: Exactly. It's the same logic as buying an oversold stock, but you've hedged away the market risk. That gives you more staying power. If the broader market drops two percent, your short leg profits partially offset the long leg loss. The spread is all that matters. Luna: Let's talk about a concrete pair right now. Do you have one you're watching? Lucas: I've been watching two regional banks that both trade in the Russell. One is a Texas bank that got hammered on a commercial real estate concern two weeks ago, down twelve percent. The other is a Southeast bank with similar exposure but only down three percent. Their ninety-day correlation is zero point seven five, but the ten-day correlation dropped to zero point three. The ratio is near a six-month high for the Southeast bank relative to the Texas one. I think the Texas bank is oversold on sentiment, not fundamentals. Luna: So you'd short the Southeast bank and go long the Texas one? That's a bet that the spread narrows as the fear fades. Lucas: Exactly. I'd enter when the ratio touches the upper band, with a stop at one and a half standard deviations beyond. The target is the mean of the ratio over the past sixty days. That gives about a three to four percent move in the spread. Luna: And how do you size it? Do you make the dollar amounts equal? Lucas: Not exactly. I size each leg so that the dollar beta to the Russell is equal. That means if the Texas bank has a beta of one point one and the Southeast bank has a beta of zero point nine, I put more dollars into the lower-beta stock to balance the market exposure. Then the pair is truly market neutral. Luna: It's a lot of moving parts, but it's a cleaner way to trade than just going long or short outright. Lucas: Absolutely. And in a market where the VIX is low and stocks are diverging, it's one of the few strategies that can generate consistent returns without betting on direction. Luna: What about the opposite scenario? If the VIX spikes and everything becomes correlated again, pairs trading falls apart. Lucas: That's the Achilles' heel. If we get a VIX above twenty-five, correlations across the board converge toward one. The spread stops mean-reverting and just moves with the panic. In that environment, you have to sit on your hands. But right now, with the VIX at fifteen, we're in a sweet spot for relative value trades. Luna: And the Russell is up, but not euphorically. There's room for dispersion to continue. Lucas: Exactly. The Russell is at two thousand nine hundred nineteen, still well below its all-time high. There's no sign of a blow-off top. Individual stocks are still being judged on their own merit. That's the perfect environment for pairs trading. Luna: One last question: how long do you typically hold a pair? Lucas: Anywhere from two days to three weeks. If the spread hasn't converged within three weeks, I re-evaluate. Often, the catalyst was weaker than I thought. Better to take a small loss and move on than to hold hoping. Luna: Fair. And if it works, you have a repeatable process, not just a lucky pick. Lucas: That's the whole point. Pairs trading is about process, not prediction. You're not guessing which stock will go up; you're betting that a temporary disconnection will reconnect.