SesTrading · Answers

Gold forecast: what history says, and what it cannot

By Saad El Soussi, UAE-authorized finfluencer · Updated 2026-11-08

Nobody can tell you where gold will be in a year, and a single price target is a guess. What history gives is a range. In every 12-month window from January 1975 to September 2026 (609 windows, monthly average gold price), gold rose in 59% of them; the middle result was +4.5%, the worst 10% were worse than -14.4% and the best 10% better than +33.3%. The worst year fell 36.4%, the best rose 197.4%. Since 2001 the picture was kinder (rose in 75%), but that is one era, not a promise.

Why a gold price target is a guess

A forecast such as "gold will be $X by December" needs a view on interest rates, the dollar, central-bank buying, wars and inflation at the same time, and a view on how the market has already priced each of them. Published targets differ by thousands of dollars for the same date. The honest alternatives are to measure how big the moves have been and to plan for several outcomes, which is what the rest of this page does.

What gold did over 1, 3, 12 and 60 months

For every month from January 1975 we compared the price with the price 1, 3, 12 and 60 months later. Windows overlap, so the counts are not independent observations; read them as a map of the range, not as a probability.

Horizon	Windows	Gold rose	Worst 10% were below	Middle result	Best 10% were above
1 month	620	306 (49.4%)	-4.2%	0.0%	+5.7%
3 months	618	328 (53.1%)	-7.6%	+0.8%	+12.7%
12 months	609	361 (59.3%)	-14.4%	+4.5%	+33.3%
60 months	561	376 (67.0%)	-25.6%	+23.5%	+156.8%

Over one month gold is a coin flip: it rose in 49.4% of months. Anyone who claims to know next month’s direction is claiming an edge this table does not show.
Over 12 months the odds tilt up, but the spread is wide. One year in ten was worse than -14.4% and one in ten better than +33.3%.
The extremes are large. The worst 12 months in the data ran from July 1980 ($643) to July 1981 ($409), -36.4%. The best ran from January 1979 ($227) to January 1980 ($675), +197.4%.
Even over five years gold fell in a third of the windows (185 of 561), and the worst 10% lost more than a quarter. By decade of the start month, 64 of 120 five-year windows starting in the 1980s and 70 of 120 starting in the 1990s ended lower, against none of 120 in the 2000s and 46 of 120 in the 2010s.

Does the 21st century tell a different story?

Yes, and it is worth seeing why not to rely on it. From January 2001 on (297 twelve-month windows) gold rose in 222 of them (74.7%); the middle result was +12.0%, the worst 10% were below -7.3% and the best 10% above +35.5%. The worst window ended in December 2013 (-27.5%) and the best in January 2026 (+75.4%). That is a long bull market with one deep drawdown in it, not a rule. The 1980s and 1990s, in the decade counts above, are what a period without it looks like.

After a strong year, what came next?

A common question is whether a big gain means a fall is due or the rally continues. In the data, in the 142 months where gold stood 20% or more above its level 12 months earlier, the following 12 months were positive in 105 (73.9%), with a middle result of +17.8%, a worst of -36.4% and a best of +197.4%. Since 2001 it was 72 of 88 (81.8%), middle +17.7%, worst -9.3%, best +75.4%. This is tempting, but these months come in a handful of long episodes (1977-1980, 1983-1987, 2002-2012, 2019-2020, 2024-2025), so it is a few stories repeated, not 142 independent tests. Gold has often kept rising after a strong year, but the worst follow-up in the table, -36.4%, also came after one: the 1979-1980 surge.

Where gold stands in the data now

The September 2026 monthly average was $4,319. A year earlier (September 2025) it was $3,668, so the 12 months to September 2026 were +17.7%. The highest monthly average of 2026 so far was $5,020 in February, so September was 14.0% below that peak. For the live price see gold price today.

To make the range concrete: if the next 12 months simply repeated the middle 80% of past 12-month outcomes (-14.4% to +33.3%), a $4,319 start would end between about $3,700 and $5,760. That is $2,000 wide. It is an illustration of the spread of the past, not a forecast, and 1 in 10 past outcomes fell outside it on each side.

How to use this as a trader

Think in scenarios, not a target. Write down what you would do if gold is up 10%, flat, or down 10% in your trading period.
Match the plan to the horizon. The one-month coin flip means a short-term trade needs a stop loss and an entry reason, not an annual forecast. Start with how to trade gold and our gold trading strategy tests.
Size from the worst case. If a bad year is -14% to -36%, a position that a 10% fall would wipe out is too big. Use the lot size calculator and read stop loss.
Check the claim, not the confidence. A forecaster who states a single number for a date has not told you the range; ask what happens if they are wrong by $500.
Keep a journal. Compare your own predictions with what happened, as in backtesting.

Limits: monthly averages smooth out the highs and lows inside a month, so real peaks and troughs were more extreme; prices are in nominal dollars, not adjusted for inflation; gold in your own currency moved differently; the windows overlap; the past is one sample of how the world worked. Run date: 6 November 2026.

Key points

No one can forecast gold reliably; a range is more honest than a target.
Since 1975, gold rose in 49.4% of months and 59.3% of 12-month windows; the middle 12-month result was +4.5%.
The worst 12 months were -36.4% (July 1980 to July 1981), the best +197.4% (January 1979 to January 1980).
Since 2001 gold rose in 74.7% of 12-month windows, but that is one era.
September 2026 average: $4,319, +17.7% in 12 months and 14.0% below the February 2026 peak month.
Plan for several outcomes and size positions from the stop; trading is risky and most retail accounts lose money.

Prefer to see gold trades with a plan?

The Telegram bot answers in English, explains the VIP gold trades (each with entry, stop loss and targets) and shows where the weekly results are published, losing trades included. Trading is risky; most retail accounts lose money.

Open the SesTrading botOn Telegram

Questions people ask

What will the gold price be in a year?

Nobody knows. Since 1975 the middle 12-month change was +4.5%, and the middle 80% of outcomes ranged from -14.4% to +33.3%. From $4,319 (September 2026 average) that range is about $3,700 to $5,760. It illustrates the past spread, not a prediction.

Will gold go up next year?

Gold rose in 59.3% of 12-month windows since 1975 (74.7% since 2001), so a rise has been more common than a fall, but about two in five years since 1975 were down and the worst fell 36.4%.

How accurate are gold price predictions?

We have not measured published forecasts, so we do not claim a number. What we can show is that gold’s one-month direction was a coin flip (up 49.4% of months) and that 12-month changes ranged from -36.4% to +197.4%, which makes single-number targets fragile.

Is gold a good investment right now?

This page cannot say. At the September 2026 average of $4,319, gold was 14.0% below its February 2026 peak month and 17.7% above a year earlier. Whether that is a good entry depends on your horizon and how much loss you can take. This is not advice.

Has gold ever fallen for years in a row?

Yes. In the monthly data the 12 months from July 1980 to July 1981 fell 36.4%, and more than half of the five-year windows starting in the 1980s and 1990s ended lower than they began. Since 2001 the worst 12-month window was -27.5% (ending December 2013).

Authorized in the UAESaad Nizar El Soussi, founder of SesTrading, is listed by the UAE Capital Market Authority as an authorized finfluencer (entry 193, added 24 September 2026). Check the official list.