PROCESS BEATS PREDICTION | BUDGET WEEK

the people who make money from markets mostly gave up guessing a long time ago

Did you see how many Budget previews landed this weekend? Everyone has a forecast for Wednesday, from income tax to whether the pound gets a kicking.

I spent years on a broking desk listening to forecasts all day. Clients loved them. I can't remember many that made anyone money (tbh, most were forgotten by lunch).

I keep coming back to a few numbers. None of them needed anyone to know what happens next.

S&P publishes a scorecard twice a year called SPIVA. It counts how many actively managed funds beat their benchmark after fees. The latest US one, the Mid-Year 2026 scorecard with data to the 30th of June, found that 90 per cent of US large-cap funds trailed the S&P 500 over 15 years. Over 20 years it was 93 per cent.

The UK version isn't any kinder. Over ten years, 91 per cent of sterling UK equity funds lagged their benchmark. Fewer than half of the UK equity funds that were running ten years ago still exist.

These are people with research teams and Bloomberg terminals. Calling the market is literally their job and most still can't beat a tracker.

AQR, a big US quant firm, went back to 1880 to test one boring idea. Buy the markets that have been going up and sell the ones that have been going down. No view on the economy. Just follow the trend across 67 markets, from commodities to currencies.

The paper is called A Century of Evidence on Trend-Following Investing. In their simulation, the strategy made money in every decade from the 1880s to the 2010s, after estimated trading costs and even after hedge fund-style fees of 2 and 20. Over the full period, that came to 7.3 per cent a year above cash, after those fees. That stretch includes two world wars and the Great Depression.

It's a backtest, so take it with a pinch of salt. You can check real money too, though.

In 2008 the S&P 500 lost 37 per cent including dividends. The Barclay CTA Index, a broad index of managed futures funds (plenty of them trend followers), made 14.1 per cent that year. Then in 2022 it happened again. The S&P lost 18.1 per cent and the SG Trend Index, which tracks ten of the biggest trend-following funds, made 27.3 per cent.

None of those funds called the sell-off. They had rules for what to do once prices started moving and they stuck to them.

This one stings a bit. We ran a backtest on Microsoft, Nvidia and Alphabet, the listed names in the MANGO basket. $10,000 split equally at the end of 2019, run to Friday's close on the 23rd of October.

Version one only bought each stock when it touched its 200-week moving average. Sounds sensible, right? You wait for a proper pullback instead of chasing.

That turned $10k into about $74k.

Version two bought all three on day one and did nothing. That got you about $148k. (For reference, the S&P 500 turned the same $10k into about $26k.)

So the patient version made half as much. Most of the gap is Nvidia. It didn't touch its 200-week average until September 2022 and by then it was more than double its price at the end of 2019. That means the dip buyer paid more than someone who bought on day one.

Microsoft was annoying in a different way. It didn't close at its 200-week average until March 2026, so that third of the money sat in cash for six years waiting.

We tried a softer version too. Hold the $10k and add $500 every week one of the three touched the line. Those top-ups came to $11,500 in total and grew to about $69,500, which looks great on its own. But putting the same $21,500 in on day one would have finished roughly $101k higher.

To be fair to the dip buyers, this is hindsight. MANGO only got its name in 2025, after everyone already knew these were the winners. Pick three different stocks and waiting for the dip might look much better.

That's sort of my point, though. Waiting for the perfect price is a prediction too. You're betting the price comes back to you. Sometimes it does. With Microsoft, it took six years.

I reckon most people reading this think they don't make predictions. Cash ISA, pay off the house, the state pension and hope the country sorts itself out.

That is a forecast. It's a bet that inflation behaves and the state can keep paying what it has promised.

Look at what the state is paying now. Treasury figures published in July put public sector debt interest for 2025-26 at £130.3bn. That's more than the whole education budget (£125.7bn) and about double defence (£65.4bn). Back in 2020-21, debt interest was £42.3bn, which was less than we spent on defence (£44.6bn).

The Budget is on Wednesday. Whatever gets announced, that interest bill is sitting on the table. If your whole plan depends on Britain working its way out of that, you've made a big call on the UK. You just never wrote it down.

I'm not telling you to buy a trend-following fund or to never buy a dip. None of this is advice.

What the funds that came through 2008 and 2022 had in common is that they decided what they'd do before the market did anything. They didn't know those years were coming. They had a plan for when prices moved and they followed it.

Be honest with yourself for a second. If the market dropped 20 per cent next month, do you know exactly what you'd do with your money? Most people I speak to don't. That's fine. It just means the plan is still a guess.

If you want to see what level your own process is at right now, the Process Score quiz will show you.

The Process Score quiz takes a few minutes and shows you what level your process is at right now.

David
Founder, FINK

P.S. If you'd rather not wait and want FINK OS now, code FINK10 takes 10% off the course (£194 off the £1,940 price) until Budget day on Wednesday the 28th of October. It closes on Wednesday night, so get FINK OS with code FINK10 here if you want it.

Use code FINK10 at checkout. Closes Budget day, Wednesday 28 October.

SOURCES

1. S&P Dow Jones Indices, SPIVA U.S. Mid-Year 2026 Scorecard (data to 30 June 2026), Report 1a: https://www.spglobal.com/spdji/en/documents/spiva/spiva-us-mid-year-2026.pdf

2. S&P Dow Jones Indices, SPIVA Europe Mid-Year 2026 Scorecard (data to 30 June 2026), Reports 1a and 2: https://www.spglobal.com/spdji/en/documents/spiva/spiva-europe-mid-year-2026.pdf

3. Hurst, Ooi and Pedersen, "A Century of Evidence on Trend-Following Investing", Journal of Portfolio Management, Fall 2017, Exhibit 1: https://www.aqr.com/-/media/AQR/Documents/Insights/Journal-Article/AQR-JPM-Fall-2017.pdf (SSRN: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2993026)

4. BarclayHedge, Barclay CTA Index annual returns: https://portal.barclayhedge.com/cgi-bin/indices/displayCtaIndex.cgi?indexCat=Barclay-CTA-Indices&indexName=Barclay-CTA-Index

5. SG Prime Services, 2022 CTA Index Performance Review (SG CTA Index and SG Trend Index), via AlphaWeek: https://alpha-week.com/2022-cta-index-performance-review

6. S&P 500 total returns by year (2008 -37.00 per cent, 2022 -18.11 per cent): https://www.slickcharts.com/sp500/returns

7. FINK backtest, Yahoo Finance adjusted closes (dividends reinvested), 31 Dec 2019 to 23 Oct 2026. MSFT, NVDA, GOOGL equal weight. Hypothetical and built with hindsight.

8. HM Treasury, Public Expenditure Statistical Analyses 2026 (CP 1624, 16 July 2026), Table 4.2: https://assets.publishing.service.gov.uk/media/6a5771e2822fbe6b8245c99f/PESA_2026.pdf

Educational only, not financial advice. Past performance isn't a guide to the future. Capital is at risk.