LEGO vs. Pokémon Investing: An Honest Comparison

If you've spent any time around alternative collectibles, you've heard the same headline about LEGO that you've heard about Pokémon: it beats the S&P 500. The numbers do, in fact, hold up well enough to take seriously. But anyone framing them as the same asset class is doing you a disservice. They look similar — nostalgia-driven physical collectibles with scarcity premiums — and they behave very differently. The structural differences are the whole game, and understanding them is the difference between buying a second copy of the same risk versus actually diversifying.

This piece is the honest comparison: where the data agrees, where it doesn't, and the five structural differences that should actually drive how (and whether) you allocate to each.

What the data actually says about LEGO

The famous number is 11% annualized, gross of fees, from a 2018 paper by Victoria Dobrynskaya and Julia Kishilova that analyzed 2,322 sets from 1987–2015. Real returns came in at 8% after inflation. The study found LEGO outperforming the S&P 500, bonds, gold, and several alternative asset classes over the sample period. Volatility was meaningful — a 28% standard deviation, a Sharpe ratio of 0.4, but positive skewness, meaning the upside tail was bigger than the downside tail. Roughly one in ten sets in the sample actually lost money. Independent follow-up work, including a BrickNerd analysis of 826 sets that retired in 2020–2021, came in at 11.7% — close enough to the original number to confirm the order of magnitude.

That's the version you'll see quoted everywhere. Here's the one you won't.

In 2020, Savva Shanaev published a paper in the Journal of Risk Finance that explicitly set out to address the survivorship-bias limitations in the earlier work. Different methodology, longer sample (1966–2018), and a much harder look at what's actually in any given index. His headline number: 1.20% inflation-adjusted return per annum. That's not a typo. Same asset class, broader sample, more careful accounting — and the return collapses to barely above zero in real terms.

Both studies can be right. The reconciliation is that certain LEGO sets — popular themes, large piece counts, exclusives, post-retirement appreciation — produce the kind of returns the famous study is measuring. The broader universe of all LEGO sets ever made produces something much closer to "it holds its value, plus a small premium." This is the LEGO version of the recency-bias problem Pokémon has. The headline number is true for a subset of the market that the analysis happens to be measuring, not for "LEGO" the way the listicles imply.

If you want the practical takeaway: well-selected LEGO investments (specifically retired exclusives in proven subthemes — more on which ones in a minute) plausibly do something in the 8–17% annualized range, depending on the basket. The "11%" you see everywhere is the right ballpark for that curated version. The "blindly buy any LEGO set" version is much closer to 1–2% real, and you should treat anyone who tells you otherwise as either uninformed or selling.

The two LEGO subthemes that actually carry the returns

Aggregate numbers hide the spread, and the spread is enormous.

Modular Buildings is the standout. Released roughly once a year since 2007 under what's now the LEGO Icons line, the modulars are big, detailed corner-and-street buildings designed to slot next to each other. The community treats them as a connected, continuous set. Retired modulars have averaged ~17% annualized growth, with 14 of 16 retired sets clearing 12%. Café Corner, the first one, went from $140 at release in 2007 to roughly $3,000 by 2015 — a 2,230% total return. That's not a typical modular outcome; it's the high end. But the broader pattern is unusually consistent.

Ultimate Collector Series Star Wars is the other reliable lane. Across roughly 1,000 retired Star Wars sets, the theme has averaged about 11% annualized. Narrow to the $300+ UCS and Master Builder Series tier specifically and that lifts to ~12%. The original 2007 UCS Millennium Falcon famously cleared $5,000 on a $500 retail price. Gift-with-purchase (GWP) promos that ship alongside major launches frequently flip for $150–$200 within weeks, providing a separate short-horizon return on top of the main set's long hold.

LEGO Ideas (community-submitted sets) is a third category worth watching but with more dispersion — bigger winners and bigger flops than the modulars.

Themes that look hot but historically underdeliver: most Friends/City, most licensed cash-grabs without a strong AFOL (Adult Fan of LEGO) following, and large playsets aimed at kids rather than collectors. Just because something has a license doesn't make it an investment.

How this compares to what we know about Pokémon

The Pokémon numbers, from the prior post: credible analyses put sealed booster boxes at roughly 21% annualized in the 2015–2020 pre-mania window, and averages in the 40%+ range in samples that start in early 2020 (i.e., that include the pandemic boom). Card Ladder's broader index, which captures all cards rather than just sealed boxes, shows 3,261% over 20 years against the S&P's 483%.

The honest read, putting both side by side: a curated, vintage-weighted, sealed Pokémon basket has produced higher historical returns than a curated LEGO basket over comparable windows. That much the data supports. But the headline gap shrinks once you account for methodology — and shrinks further once you account for the structural differences in how those returns are produced. Which is the actual point.

The five differences that should drive how you allocate

1. Supply discipline. This is the single biggest structural difference, and it's not close.

LEGO retires sets on a publicly knowable schedule — typically 2–3 years for most sets, 3–5 for major UCS builds. When a set retires, production stops. Permanently. LEGO can release a new set in the same theme, but it cannot reissue the same SKU. This is announced policy, which means the supply curve becomes fixed on a date investors can plan around.

The Pokémon Company prints on a fundamentally different scale. Roughly 12 billion cards in a recent fiscal year. Sets that move well get reprinted. Popular product runs longer. Modern Pokémon supply is, structurally, "as much as the market will absorb." Vintage Pokémon supply is fixed in the same way LEGO supply is fixed, but the modern product flowing through the market today does not have LEGO's supply discipline. That's a real structural disadvantage for modern Pokémon as an investment vehicle versus modern LEGO. Vintage Pokémon vs. retired LEGO is a much more even matchup.

2. What you're actually buying.

A retired LEGO set is bricks, instructions, and minifigures. If you build it, you get a display piece. If the investment thesis doesn't pan out, you still have a thing. There is a utility floor under the asset.

A sealed booster box is a box of cards whose entire investment premium evaporates the moment you open it. There is no utility floor — sealed value and opened value are two completely different numbers, and the gap is huge. This makes Pokémon sealed product closer to a pure financial instrument and LEGO closer to a hybrid collectible. For some investors that's a feature (no temptation to open it ruins your hold); for others it's a bug (you're holding an envelope you literally cannot peek inside without destroying the value).

3. Volatility and cycle exposure.

Dobrynskaya's data found LEGO has "modest positive exposure" to equity markets and "low crash risk" — translation: it doesn't move violently with stocks and doesn't have a fat downside tail. Behavior closer to a slow-cooker collectible.

Pokémon is the opposite. The 2020–2021 boom was historic; 2022 saw real corrections; cycles are tied to media events (anniversaries, anime drops, viral pack-opening moments). It's a more momentum-driven, cycle-exposed asset. Higher peak returns come with more cycle risk and more frequent drawdowns.

If your existing portfolio is already equities-heavy and you're looking for genuine diversification, LEGO's lower correlation and lower volatility is doing more diversification work than Pokémon's. If you're looking for higher peak returns and you can tolerate the cycle, Pokémon has historically produced them.

4. Capital efficiency and storage burden.

This one is purely practical and rarely discussed. A serious sealed-Pokémon portfolio fits in a closet. A serious LEGO portfolio is a storage problem. UCS sets are physically enormous — the new Death Star and the larger modulars take up significant shelf real estate even sealed. Climate-controlled storage isn't optional past a certain portfolio size.

Said differently: per dollar invested, LEGO takes meaningfully more physical space than Pokémon does. For a small allocation it doesn't matter. For a serious one, it's a real friction that compounds — insurance, rented storage, and the slow accumulation of cubic feet that have to live somewhere.

5. Counterfeit and condition risk.

LEGO has condition risk — corner crushing, sticker yellowing, plastic deterioration on older sealed sets — but largely doesn't have a counterfeit problem. Almost nobody is forging entire sealed LEGO boxes; the unit economics don't support it.

Pokémon has both. Resealing scams (open, take chase cards, professionally reseal, sell as "sealed") are a real and documented industry problem, and they're getting more sophisticated. Counterfeit modern singles are advanced enough to fool inexperienced graders. The infrastructure of authentication and trusted sellers in Pokémon is more important — and a bigger friction — than it is in LEGO.

Where each one actually wins

LEGO wins on:

• Predictability — retirement schedules are public, supply discipline is policy
• Lower volatility and genuine portfolio diversification (low equity correlation)
• A utility floor — the asset is something even if it doesn't appreciate
• Lower counterfeit risk and a more transparent secondary market

Pokémon wins on:

• Capital efficiency — much more value per cubic foot of storage
• Faster cycles — modern sets often show meaningful appreciation within 1–2 years rather than waiting on a 3–5-year retirement clock
• Potentially higher peak returns in demand-cycle years (the 2020–21 numbers nobody got close to)
• Lower friction per transaction (booster boxes ship in a small flat box)

So how should this actually inform a portfolio?

The honest framing isn't "LEGO vs. Pokémon, pick one." It's that they're two different tools that happen to share a category name. A serious alternative-collectible allocation that already includes Pokémon arguably benefits more from LEGO than from doubling down on Pokémon, because LEGO's lower correlation and supply discipline are doing different work than another year of sealed booster boxes does. The diversification case for adding LEGO to a Pokémon-heavy portfolio is stronger than the case for adding more Pokémon to it.

But — and this is the same point as the prior post — neither one is a substitute for diversified, productive-asset investing. Both are alternative slices. The Dobrynskaya number that survives scrutiny is closer to 8% real than 11% gross. The Pokémon number that survives the recency-bias adjustment is closer to 20% than 40%. Both still beat the long-run S&P 500 number, and both come with frictions, illiquidity, and risks that traditional investments don't have. Treat them as a small, eyes-open slice of a portfolio, not a thesis you stake the savings on.

If you take one thing from this piece: the structural differences between these two assets matter more than the headline return gap. Modular Buildings are not a sleeker version of Evolving Skies. They're a different bet on a different timeline with a different risk profile — and that's exactly what makes them potentially worth holding alongside each other.

This post is documentation and analysis of an alternative asset class. It is not financial advice, and no part of it should be treated as a recommendation to buy, sell, or hold any specific item. Do your own diligence; use your own judgment.