Equity research · 2016 to 2025
Beating the S&P 100% of the time.
Ten years. Ten portfolios. Every single one beat the index, by 53% in the worst year and 198% in the best. There is one small catch: we picked them after the year had ended. So the real question is what the winners had in common, and whether you could have spotted them in January.
Hypothetical portfolios built with hindsight · ten stocks a year, one per industry, equal weights, held January to December · benchmark SPY
The headline
Ten out of ten years, the winners left the index behind
Each January we put an equal amount into ten of that year’s best-performing large companies, one per industry, and held them to December. Here is how each basket did against SPY, the fund that tracks the S&P 500.
10 / 10
years the winners’ basket beat SPY
+198%
biggest lead, 2020: +213% against +15%
+53%
smallest lead, 2018: +46% while SPY fell 7%
12 / 12
months ahead of SPY in both 2020 and 2025
| Year | 10-stock basket | SPY | Lead | Months ahead |
|---|---|---|---|---|
| 2016 | +113% | +11% | +102% | 10 / 12 |
| 2017 | +111% | +18% | +93% | 10 / 12 |
| 2018 | +46% | −7% | +53% | 10 / 12 |
| 2019 | +109% | +29% | +80% | 11 / 12 |
| 2020 | +213% | +15% | +198% | 12 / 12 |
| 2021 | +115% | +29% | +86% | 10 / 12 |
| 2022 | +53% | −20% | +73% | 10 / 12 |
| 2023 | +122% | +25% | +97% | 9 / 12 |
| 2024 | +159% | +24% | +135% | 10 / 12 |
| 2025 | +148% | +17% | +131% | 12 / 12 |
Lead is the basket’s return minus SPY’s. Months ahead counts the calendar months in which the basket did better than SPY. A pale bar means SPY fell that year.
The surprise
It wasn’t just Nvidia and Tesla. It was also mustard.
The winners included oil producers, banks, hotels, insurers, retailers, drug makers, cruise lines, a hospital chain, a used-car dealer and McCormick, the company behind French’s mustard. Big technology names showed up, but they were never the whole story.
So we went through all 100 picks one by one: the business, the industry it sat in, what it cost at the start of the year, and above all, the event that made the shares take off. The businesses looked nothing alike. The reasons they won looked remarkably similar.
Year by year
Ten years, ten very different winners
Tap a year for the story behind it and the ten stocks in the basket.
2016When the market stopped expecting disasterBasket +113% · SPY +11% · lead +102% · read
AMD started the year near USD 2 a share, losing money and written off against Intel. Then came a licensing deal, well-reviewed graphics cards and the first look at its Zen processors. It ended the year up 309%, still loss-making, but now priced as a company with a future.
Much of the rest of the basket was the same story in other industries: Vale and Petrobras on recovering commodity prices, Santander Brasil on a stronger real. Depressed businesses, then the conditions changed.
The opportunity wasn’t the best business. It was a business the market had written off too early.
2017When strong growth became even strongerBasket +111% · SPY +18% · lead +93% · read
This time the market paid for businesses that were already growing, and kept beating expectations. China Lodging (+187%) franchised hotels across a fast-growing travel market; its earnings multiple went from 28 to 60. Arista Networks (+143%) sold the switches inside cloud data centres and kept taking share from Cisco.
Nvidia (+90%) was already riding data-centre demand, years before AI became a household word.
The market wasn’t paying for growth. It was paying for growth that kept surprising on the upside.
2018A good year for mustard and car partsBasket +46% · SPY −7% · lead +53% · read
SPY fell 7%. The basket made 46%. The Trade Desk (+150%) beat forecasts every quarter as advertising moved to connected TV. But the rest was gloriously ordinary: O’Reilly Automotive (+38%) selling car parts to garages that can’t wait for Amazon, and McCormick (+38%), whose mustard and hot-sauce acquisition the market expected to disappoint. It didn’t.
Nobody needed to invent artificial intelligence to make this work.
2019The companies that rewrote their own storiesBasket +109% · SPY +29% · lead +80% · read
Shopify (+189%) went from online-shop software to infrastructure for retailers. AMD (+144%) came back, this time because its chips were genuinely ahead of Intel’s. And Target (+93%), supposedly a casualty of Amazon, turned its stores into a delivery and pick-up network and jumped 20% in a day on the results.
The common theme was a change in what the market thought the business could become.
2020When the future arrived earlyBasket +213% · SPY +15% · lead +198% · read
The pandemic pulled years of change into months. Tesla (+720%) delivered profit after profit and won a place in the S&P 500. Sea Limited (+397%) and Pinduoduo (+331%) rode the rush to online shopping. Roku, Nvidia, a copper miner and a pet-diagnostics company joined them.
The basket beat SPY in all twelve months. Even in the March crash it fell less: −12.3% against −13.0%.
This wasn’t one lucky stock. It was a group of companies the market had underestimated all year.
2021Vaccines, hackers, pickup trucks and shopping mallsBasket +115% · SPY +29% · lead +86% · read
BioNTech (+201%) saw vaccine sales forecasts nearly double in eight months. Fortinet (+147%) sold protection against the ransomware wave. Ford (+144%) took 200,000 reservations for its electric pickup. And Simon Property (+93%), the shopping-mall owner, won on the exact opposite trend: people going back out.
Finding the winners didn’t require one economic story. It required spotting which companies faced the biggest positive change.
2022The year the market rediscovered profitsBasket +53% · SPY −20% · lead +73% · read
SPY fell 20%. The basket rose 53%. Inflation, rate rises and the war in Ukraine rewarded companies with cash today. Occidental Petroleum (+103%) paid down debt while Berkshire Hathaway kept buying. Marathon Petroleum (+77%) saw refining margins triple. Defence, pharma and healthcare names filled the rest.
The question wasn’t oil or technology. It was which businesses were about to earn far more than expected.
2023The year the earnings forecasts brokeBasket +122% · SPY +25% · lead +97% · read
On 24 May Nvidia (+246%) forecast USD 11bn of quarterly revenue when analysts expected 7.2bn. The shares jumped 25% in a day. Meta (+184%) declared a ‘year of efficiency’ and its earnings multiple went from 12 to 31. Nubank’s parent, a mortgage lender and Ryanair made the list too.
AI dominated the headlines. The connecting theme was the size of the positive surprise.
2024AI, Bitcoin, police cameras and cruise shipsBasket +159% · SPY +24% · lead +135% · read
Palantir (+356%) saw commercial demand for its AI platform surge; investors went from paying 18 times sales to 70. MicroStrategy (+323%) became a leveraged Bitcoin play. Nvidia (+179%) grew profits so fast its shares got cheaper on earnings while rising. Royal Caribbean, Axon and Howmet’s jet-engine parts filled out a very odd-looking basket.
November alone delivered 43% of the year’s lead.
Even extraordinary businesses don’t reward investors on a smooth schedule.
2025The gold rush moves to the suppliersBasket +148% · SPY +17% · lead +131% · read
AI spending reached the companies selling the picks and shovels. Micron (+227%) and Seagate (+219%), memory chips and hard drives, rode a supply squeeze from data centres. Robinhood, Banco Santander, a gold miner and a used-car dealer joined them.
Like 2020, the basket beat SPY in every one of the twelve months.
Several genuinely good ideas can build a remarkable lead when their catalysts keep developing.
What the winners shared
Four things worth looking for
Different industries, different valuations, different sizes. Yet the same four features kept turning up.
01
The market didn’t see it coming.
A good business isn’t enough. You need one whose future turns out better than the price assumes. AMD in 2016, Target in 2019, Meta in 2023: each was priced for disappointment, and each proved the market wrong.
02
A real tailwind behind the industry.
Oil and iron ore recovering in 2016, the pandemic in 2020, the energy shock in 2022, AI spending from 2023. But the tailwind only matters if the company is unusually well placed to catch it.
03
Results that keep beating forecasts.
Most winners didn’t rally once. They beat, raised guidance, and beat again: Arista, The Trade Desk, Palantir, Micron. A company growing 15% when everyone expected a decline can do better than one growing 30% when everyone expected 50%.
04
Price is about tomorrow, not today.
Some winners started cheap, some started expensive, and some got cheaper while their shares soared. A low price-to-earnings ratio would have missed half of them. The question is where earnings are heading against what the market already expects.
The valuation question
Expensive can get dearer. Cheap can stay cheap.
A valuation screen tells you what the market is paying. It doesn’t tell you whether the market is right.
The path matters
The lead built month by month, not in one lucky trade
In 2020 and 2025 the basket beat SPY in every single month. In most other years it was ahead in at least ten. In 2025 the lead was 18% by March, 59% by June, 110% by September and 131% by December.
It was never smooth, though. In January 2016 the basket fell 12.4% while SPY lost 3.6%. Anyone who gave up after four weeks missed a 29.6% gain in March. And in 2024, a single month, November, delivered 43% of the year’s lead.
The market rarely sends an invitation before the best month. The answer isn’t blind patience. It’s knowing why you own the company, what supports the thesis, and what would prove it wrong.
The investor’s playbook
How to hunt for the next winner
1. Start with what could change the business. A new product taking share, a shortage that lets prices rise, a turnaround. Ask what could happen in the next 6 to 18 months that changes its earnings power.
2. Find who is best placed to benefit. When AI spending took off, Nvidia was obvious. Memory, storage and fibre makers came later. The biggest winner is not always the one getting the most attention.
3. Follow the numbers, not the narrative. Orders, market share, prices, margins, guidance. A company that keeps beating expectations deserves more attention than one whose story depends on 2030.
4. Compare the opportunity with what’s priced in. The question isn’t whether the company is good. It’s whether it’s likely to be better than the market expects.
Five things to check
| Industry demand | A change large enough to improve the company’s economics |
| Competitive position | Evidence the company can capture an outsized share of it |
| Earnings revisions | Results and guidance that keep beating expectations |
| Valuation | A price that doesn’t yet reflect the better outlook |
| The thesis | Clear milestones over the next 6 to 18 months, and the reasons it could fail |
The bottom line
The next great stock probably looks ordinary today
In Back to the Future Part II, Biff gets hold of a sports almanac from the future and bets on every result. Picking the best stocks after the year has ended is the same trick. It works brilliantly, provided somebody invents a time machine.
The point isn’t that finding the winners is easy. It’s that the gap between the market average and the exceptional company is enormous, and that the winners kept coming from recognisable changes in real businesses: a product doing better than expected, an industry coming out of a slump, a turnaround, an earnings outlook revised up again and again. Those changes showed up in announcements, results and guidance. The hard part is seeing what they mean before everyone else does. That takes research, judgement, patience and some luck.
You don’t need to find every future winner. You need to recognise a few before they become obvious.
How we built it
For each year from 2016 to 2025 we selected ten of the best-performing stocks in a universe of the 500 largest companies, with no more than one per industry. Each basket is equally weighted, bought on the first trading day of January and held to the end of December. The benchmark is SPY. Returns and valuation multiples are rounded; the lead is the difference between the two percentage returns. Monthly performance and approximate trailing valuation ratios come from public sources.
Important information
These are hypothetical portfolios chosen with hindsight. They are not actual managed returns or a strategy anyone could have followed in real time. No advisory fees, trading costs, taxes or dividend adjustments are included. This page is for general information and discussion only and is not investment, legal or tax advice, nor an offer or recommendation to buy, hold or sell any security. All investments involve risk, including loss of capital; individual shares, concentrated portfolios and highly valued companies can be very volatile. Past performance is not a reliable indicator of future results. Analysis as of 5 November 2026.