If I were to gather the person who has made the most accurate predictions about a financial product—such as gold—the most authoritative institutions, and the most famous analysts, and compare each of their predictions with the actual results to find out "who is the most accurate"... then look at how these "most accurate people" view the future now—
Would I then have mastered the wealth code of this financial asset?💰
With this thought, I really went ahead and did it. I took gold as a sample and analyzed the prediction records from over a decade.
For this research, we brought out three types of people: the top investment banks and industry organizations on Wall Street, the loudest voices in the gold sector, and the "proven experts" who accurately predicted key reversals.
Let’s look at the data one by one.
The prediction data we found is all laid out.
Wall Street professional institutions:
The LBMA (London Bullion Market Association) invites dozens of top analysts every year to make annual predictions on gold. In 2025, the average prediction from 28 analysts was $2,735/ounce. The most optimistic analyst that year—Keisuke (Bill) Okui from Sumitomo Corporation—gave $2,925, and because he was "closest to reality," he received that year's "most accurate prediction award."
What is the actual average gold price in 2025? $3,431.
That is to say, the most bullish analyst in the entire market, who ultimately won, still predicted a value 15% lower than the actual. The market consensus underestimated it by a full 20%.
Goldman Sachs has two very notable records in the history of gold predictions. In April 2013, Goldman issued a report clearly recommending shorting gold, with a target of $1,450. Gold subsequently plummeted 26%, and Goldman became renowned.
But recently, Goldman Sachs flipped. In October 2024, Goldman predicted the gold price for 2025 would be $2,700. What happened? The gold price soared to exceed $5,600 in early 2026. It missed by a factor of two.
JPMorgan gave a benchmark of $5,055 for gold prices at the end of 2025. As a result, gold prices broke through this level ahead of schedule.
Gold Dao Tai V:
Peter Schiff, the most famous "always bullish" player in the gold circle. He has been calling for "$5,000 gold" for over a decade. Gold prices were stagnant from 2013 to 2018 for five or six years, and he was criticized daily, mocked as a "broken clock." But by early 2026, gold prices indeed broke through $5,600. After shouting for over a decade, he finally got it right.
Jim Rogers, a legendary investor in the commodity market. In the early 2010s, he predicted gold would rise above $2,000, which was considered ridiculous at the time. Now it seems he was right about the direction, but he was ten years off on the timing.
Mike Maloney, creator of the "Currency History" video series, is a deep gold bull. He has long predicted that gold is severely undervalued and will eventually return to its true historical currency value. Between 2015 and 2020, his predictions were consistently validated by the market as being too optimistic. After 2020, when the price of gold started to rise, he began to be considered "finally correct."
The chosen ones:
Nouriel Roubini (Dr. Doom) is most famous for accurately predicting the financial crisis in 2008. Regarding gold: In 2013, when gold prices dropped from $1,900, he said to continue being bearish between $1,500 and $1,600, and gold truly broke below the $1,200 low point, perfectly validating his prediction. In January 2023, with gold prices hovering around $1,900, he turned bullish, predicting a 10% annual increase for five years, targeting $3,000. Gold prices later far exceeded this number.
Ben McMillan (Chief Investment Officer of IDX Advisors) has stood out in recent market conditions. At the beginning of 2024, when gold was around $2,000, he predicted it would reach $5,000 within five years. The market at that time thought it was "almost crazy." As a result, gold prices reached that level in just a year and a half.
Ray Dalio (founder of Bridgewater) does not give specific prices but makes qualitative judgments from a macro cycle perspective. In January 2026, he referred to gold as "the second largest currency," suggesting an allocation of 5-15%.
After reviewing the data, you might think—are some people actually quite accurate?
Don't rush. The above are just their "most famous few times." When I pulled out their complete records, the picture was different.
Wall Street professional institutions: A typical lagging prediction.
What is lagging prediction? It is when the bull market has already arrived that they start to raise target prices; but the adjustment always lags behind the actual increase. When the bear market comes, they begin to lower their targets, but they always do so too slowly.
The 28 analysts from the LBMA are the best example. They make one prediction a year, essentially making slight extrapolations of "trends that have already occurred." In 2024, gold prices have already risen to $2,700, yet their median forecast for 2025 is only $2,735—almost just bringing last year's closing price over as a prediction. As a result, the average price in 2025 was $3,431, contradicting their prediction by 20%.
Goldman Sachs follows the same pattern. At the end of 2024, for 2025, they only gave $2,700, and later gold prices surged past $5,000. JPMorgan gave a benchmark price of $5,055, and gold prices broke through ahead of schedule.
What these institutions are doing is more accurately described as **"trend confirmation"**—telling you that things that have already happened are indeed happening, but their judgment on the magnitude is always conservative. If you wait for their signals to make decisions, you will always be one step slow.
Big track V: A broken clock can be right twice a day.
Peter Schiff has been calling for $5,000 gold for over a decade. Jim Rickards has been calling for $10,000. Kiyosaki directly calls for $35,000.
Their strategy is essentially to shout bullish every year; if it goes up, it’s "I said it long ago," and if it goes down, it’s "not yet the time."
The more fatal problem is: predictions of this kind lack temporal granularity. They don't tell you when to enter or when to run. If you had listened to Schiff and fully invested in gold in 2011, you would have had to endure five or six years of sideways movement and losses before getting to today. Faith in this thing does not have a hemostatic function when you've lost 40%.
The chosen ones: Are they really always right?
This type of person is the most misleading. Because they have indeed made astonishingly accurate judgments at certain key moments, the market has given them the halo of "prophet." But when I pulled out their complete records, the picture wasn't so perfect.
Roubini was bearish in 2013 and correct, and he's bullish in 2023 and correct. He captured both turning points, which is indeed impressive.
But do you know what he missed in between? When gold prices just broke $1,000 in 2009, Roubini publicly said, "It’s impossible to rise another 20-30%." What happened? Gold prices soared to $1,900 in 2011, an increase of nearly 90%. By the end of 2009, when gold reached $1,200, he again said, "It looks very much like a bubble," and "Gold has no intrinsic value."
Throughout the entire gold bull market from 2009 to 2012, Roubini repeatedly sang bearish, completely missing out. No one talks about this part of history; everyone only remembers his beautiful bearish call in 2013 and his bullish call in 2023.
Ben McMillan predicted $5,000 within five years at the beginning of 2024, and it reached that in just a year and a half. His logic is based on structural changes in central bank gold purchases, which is indeed correct. But the problem is: this is the only widely documented prediction he has made in the gold field. The sample size is just one. Does one correct prediction indicate systemic predictive ability?
Ray Dalio sounds the most stable—he doesn’t predict prices, only gives allocation advice. But if you look at his macro prediction record: In 1981, he firmly believed that the U.S. would face a Great Depression, shouting everywhere in newspapers, on TV, and at congressional hearings, only to be completely wrong, almost causing Bridgewater to collapse, and he had to borrow $4,000 from his father to pay family bills. In 2015, he said, "It will replay 1937," which didn’t happen. In 2018, he said, "Recession within two years," which didn’t happen. In October 2022, he shouted, "perfect storm"—that month was exactly the bottom of the U.S. stock market.
He predicts a financial crisis almost every two to three years, and most of them don't happen. But ironically, his statement, "You don’t need to predict prices, just allocate 5-15%," has become the most useful phrase among everyone.
The script from 2011 is being replayed in 2026.
There is a particularly interesting discovery in the report.
Before gold prices peaked at $1,923 in 2011, market predictions crazily escalated in a stepwise manner: at the beginning of the year, everyone predicted $2,000, doubling by mid-year, and near the peak, Jim Sinclair called for $12,500, and Rob Kirby called for $15,000. The most extreme predictions appeared just weeks away from the actual peak.
Then in September, gold prices plummeted. What was the reaction of the predictors? First they said, "healthy correction," then a few months later, they reluctantly lowered their target price by 20-30%, and finally pushed the timetable indefinitely.
In March 2026, gold prices plummeted 25% from the historical high of $5,600 to around $4,200—the largest single-week drop since 1983. What were the reactions of most institutions and celebrities? They maintained their original high target prices and even considered the drop as "the best buying opportunity."
History does not simply repeat, but the script is indeed very similar.
So how do they see the future now?
Since we’ve dug it all up, let’s list their latest judgments for everyone’s reference:
Person/Institution Latest Prediction Core Logic Roubini Previously targeted $3,000 has been realized, bullish direction unchanged Inflation expectations returning + long-term structural increase McMillan $10,000 within five years Central bank gold purchases + U.S. debt crisis + BRICS de-dollarization Dalio does not give a price, suggests allocation of 5-15% Structural decline in fiat currency credit Jamie Dimon may reach $10,000 this year Economic concerns + inflation + asset bubbles Peter Schiff $11,400 within three years Claims recent declines are "illogical" Kiyosaki $35,000 After "the largest bubble burst in history" JPMorgan $6,300 believes the plunge is profit-taking Goldman Sachs $5,400 Bull market not over UBS $6,200 maintains bullish.
Did you see that? From $5,400 to $35,000, the highest and lowest differ by nearly 7 times. In the same market environment, using the same data sources, the answers given by the world's top minds can vary so much.
So, have you found the "wealth code"?
My conclusion after finishing all the sorting: I found nothing.
Institutions are always chasing, big names are always shouting, and the chosen ones aren't always right—they're just correct at certain specific moments, and no one remembers when they are wrong. Stacking predictions from these three types of people together doesn’t yield a more accurate answer but rather creates more confusion. Because they often contradict each other at the same time.
Originally, I thought that "finding the most accurate person and following him" was a way. After completing this research, I found that in the field of gold prediction, there is no "consistently most accurate person." There are only "those who happened to be correct this time."
Written at the end.
A single gold made me completely demystify so-called financial experts.
Whether ALPHA can be caught by you, besides models and data, it may really also depend on destiny.
So, in the end, rather than trying to crack the wealth code, I decided it’s better to learn from Dalio—don’t predict specific prices, acknowledge uncertainty, and manage risks through allocation.
Gold was purchased last year and will continue to be purchased this year. Individuals calculate their investment time dimension based on a 10-year cycle.