Gold at $5,000+. Central banks buying at record pace. The gold/silver ratio signaling opportunity. Precious metals are having their biggest structural moment since the 1970s. Here is how to understand and trade them.
Gold is not just a commodity — it is money. For 5,000 years, civilizations have used gold as a store of value, unit of account, and medium of exchange. Even after the abandonment of the gold standard in 1971, gold remains the ultimate backstop: when everything else fails — currencies, bonds, banks — gold holds its value.
In March 2026, gold trades above $5,125/oz, up from $2,000 just two years ago. This historic rally is driven by three converging forces: central bank buying at record levels, de-dollarization by emerging markets, and geopolitical instability (Iran/Hormuz, Ukraine, Taiwan). Understanding what drives gold is essential for any portfolio manager.
Gold has properties that no other asset possesses:
| Source | % of Total Demand | Trend | Key Insight |
|---|---|---|---|
| Jewelry | ~45% | Stable | India and China dominate. Price-sensitive (high prices reduce demand) |
| Central Banks | ~25% | Rising sharply | 1,000+ tonnes/year since 2022. De-dollarization driver. China, India, Turkey leading |
| Investment (ETFs/bars) | ~20% | Cyclical | ETF flows track real rates. Western investors add during rate cuts |
| Technology | ~8% | Stable | Electronics, dentistry, aerospace. Small but steady |
| Mining Supply | ~3,600 t/year | Flat | New discoveries declining. Existing mines depleting. Peak gold supply? |
Question: If central banks continue buying 1,000+ tonnes per year and mine supply is flat at 3,600 tonnes, what is the implication for prices?
Answer: Central banks now absorb approximately 28% of annual mine production. Combined with jewelry demand (~1,600 tonnes) and investment demand (~700+ tonnes), total demand exceeds supply by a growing margin. With no significant new mine supply coming online (new gold mines take 10-15 years to develop), the structural supply deficit should support prices. This is the fundamental bull case for gold: demand is growing structurally while supply is constrained. It is similar to the oil underinvestment thesis — but for gold, the supply constraint is even harder to resolve.
The single most important variable for gold prices is US real interest rates — the yield on Treasury bonds minus inflation. For the past 20 years, gold has had a strong inverse correlation (~-0.85) with the 10-year TIPS (Treasury Inflation-Protected Securities) yield.
Gold does not pay interest or dividends. When real rates are high (e.g., +2%), holding gold has a high opportunity cost — you could earn 2% above inflation risk-free in Treasuries. Why hold a non-yielding asset?
When real rates are low or negative (e.g., -1%), the calculus flips. Treasury bonds are losing purchasing power. Cash is being eroded by inflation. Gold's zero yield becomes attractive because everything else has a negative real yield. Gold becomes the least-bad option for preserving purchasing power.
The extreme case: in 2020, real rates hit -1.0%. Gold surged to $2,070. In 2022, real rates rose to +2.5% as the Fed hiked. Gold dropped to $1,620. In 2024-2026, as rate cuts began and inflation expectations rose, gold surged past $5,000.
Something unusual happened in 2024-2026: gold broke its correlation with real rates. Real rates remained positive (~1.5%), yet gold surged from $2,000 to $5,000+. The reason: central bank buying overwhelmed the rates signal. When central banks buy 1,000+ tonnes per year (10% of total gold stock flow), they create demand that is price-insensitive and does not respond to interest rates.
This is a structural shift. Gold now has two drivers: (1) traditional real rates correlation (still works for Western ETF flows), and (2) central bank/sovereign demand (independent of rates). The second driver is new and may be permanent as de-dollarization accelerates.
Old model: Gold = f(real rates, USD, risk sentiment). New model: Gold = f(real rates, USD, risk sentiment, central bank buying, de-dollarization). The new variables are structural and long-term. This means gold may stay elevated even if real rates rise, as long as central bank buying continues. This is why the traditional "rates up = gold down" playbook has been unreliable since 2024.
Question: The Fed cuts rates by 50 basis points. Inflation expectations remain anchored at 2.5%. The 10-year yield drops from 4.0% to 3.5%. What happens to real rates and gold?
Answer: Real rate = 3.5% - 2.5% = 1.0% (down from 1.5%). The 50 bps decline in real rates is bullish for gold. Historically, a 50 bps decline in real rates corresponds to a 5-8% gold rally. Combined with the signal that the Fed is easing (dovish), which weakens the dollar, gold should rally on both channels: lower real rates + weaker dollar. Expect a $200-400 move over the following weeks.
Central bank gold purchases have been the single most important driver of gold prices since 2022. After the US froze Russia's $300B in foreign exchange reserves following the Ukraine invasion, every central bank in the world took notice: dollar reserves can be weaponized. The response has been a dramatic shift toward gold.
| Central Bank | Net Purchases (tonnes) | Total Holdings (tonnes) | % of Reserves in Gold |
|---|---|---|---|
| People's Bank of China | ~400 (reported, likely more) | ~2,400 | ~5% (target: 10-15%) |
| Reserve Bank of India | ~250 | ~900 | ~9% |
| Central Bank of Turkey | ~200 | ~600 | ~30% |
| National Bank of Poland | ~180 | ~450 | ~15% |
| Czech National Bank | ~50 | ~80 | ~6% |
| Singapore MAS | ~75 | ~250 | ~4% |
China officially holds ~2,400 tonnes of gold (5% of reserves). But estimates suggest the real number is much higher — perhaps 4,000-5,000 tonnes — because China buys through state-owned banks and does not always report purchases immediately. If China is targeting 10-15% of reserves in gold (as analysts estimate), they need to buy another 3,000-5,000 tonnes at current reserve levels.
At current mine production of 3,600 tonnes/year globally, China alone could absorb nearly an entire year of global production. This is why the gold bull case is so powerful: the marginal buyer (central banks) has an almost unlimited appetite, and the supply response is minimal.
Question: If China increases its gold holdings from 5% to 10% of reserves, how many tonnes would they need to buy and what percentage of annual global production does that represent?
Answer: China's total reserves are approximately $3.3 trillion. At current gold price of ~$5,000/oz (~$160M/tonne), 5% = ~$165B = ~1,030 tonnes (roughly current holdings). 10% = ~$330B = ~2,060 tonnes. They need to buy approximately 1,030 additional tonnes. Global annual mine production is ~3,600 tonnes. So China's target increase represents 29% of one year's global production. Spread over 3-5 years, that is 200-350 tonnes of additional annual demand — equivalent to adding another major gold consumer to the market. This is a massive structural tailwind for gold prices.
Silver is gold's more volatile cousin. It trades like gold during monetary crises (safe haven) but also like copper during economic expansions (industrial demand). This dual nature makes silver uniquely interesting — and uniquely challenging to analyze.
| Use | % of Demand | Trend | Key Detail |
|---|---|---|---|
| Industrial | ~55% | Rising (solar) | Solar panels use 100M oz/year and growing 20%+ annually. Electronics, 5G, EVs |
| Jewelry/Silverware | ~20% | Stable | India dominates jewelry demand |
| Investment (ETFs/bars) | ~20% | Cyclical | Follows gold sentiment. SLV ETF flows important |
| Photography | ~3% | Declining | Digital killed film but niche demand remains |
Solar panels are the fastest-growing source of silver demand. Each standard solar panel uses approximately 20 grams of silver (paste for conductive contacts). Global solar installation is growing 20-30% per year, and newer technologies (heterojunction, TOPCon) use MORE silver per panel, not less.
By 2030, solar alone could consume 200-250 million ounces of silver per year — compared to total mine production of ~820 million ounces. If solar takes 25-30% of total silver supply, the industrial squeeze on silver could mirror what EVs are doing to copper. This is the structural bull case for silver: the energy transition needs silver, and mines cannot scale fast enough.
| Attribute | Gold | Silver |
|---|---|---|
| Primary Role | Monetary / store of value | Hybrid: industrial + monetary |
| Volatility | 15-20% annualized | 25-35% annualized (1.5-2x gold) |
| Beta to Gold | 1.0x | ~1.5-2.0x (amplifies gold moves) |
| Crisis Behavior | Rallies consistently | Initially sells off with equities, then catches up |
| Central Bank Buying | Massive (1,000+ tonnes/year) | None (central banks do not hold silver) |
| Storage Cost | Low (high value-to-weight) | Higher (5x less value per kg) |
| Market Size | ~$12 trillion (above ground) | ~$50 billion (tiny — easily squeezed) |
Question: Gold has rallied 20% over the past 6 months. Silver has only risen 12%. What does this divergence suggest?
Answer: Silver is underperforming relative to gold, which typically means silver will catch up with a sharp move. Historically, when silver lags gold in the early stage of a precious metals bull market, it catches up and eventually outperforms in the later stages. The gold/silver ratio is probably elevated (above 80). This sets up a "silver catch-up trade" — long silver, short gold — that benefits when the ratio compresses. Alternatively, simply go long silver for a leveraged bet on the continuation of the gold rally.
Platinum and palladium are the "other" precious metals, but their demand drivers are completely different from gold and silver. They are primarily industrial metals used in automotive catalytic converters (which clean exhaust emissions), with limited monetary demand.
| Attribute | Platinum | Palladium |
|---|---|---|
| Price (Mar 2026) | ~$1,050/oz | ~$950/oz |
| Primary Use | Diesel catalytic converters, jewelry, hydrogen fuel cells | Gasoline catalytic converters |
| Auto Demand % | ~35% | ~80% |
| EV Impact | Positive (hydrogen fuel cell demand) | Very Negative (EVs have no catalytic converter) |
| Supply Concentration | South Africa (70%), Russia (12%) | Russia (40%), South Africa (35%) |
| Investment Demand | Growing (PPLT ETF) | Declining |
| Substitution | Can substitute for palladium in gasoline converters | Being replaced by platinum (cost advantage) |
A fascinating dynamic is playing out: automakers are substituting platinum for palladium in gasoline car catalytic converters. In 2019, palladium was $2,500/oz and platinum was $900/oz — a massive price differential. Automakers responded by re-engineering catalytic converters to use platinum instead. This substitution is structural, takes 2-3 years to implement, and is now well underway.
The result: palladium demand is falling (EV transition + platinum substitution), while platinum demand is rising (substitution + hydrogen fuel cell potential). This is a classic convergence trade — long platinum, short palladium.
Question: EVs have no catalytic converters. If EV adoption reaches 50% of new car sales globally, what is the impact on palladium demand?
Answer: At 50% EV penetration, palladium demand from the auto sector (which is 80% of total demand) would fall by approximately 40% (50% fewer ICE cars x 80% auto demand share). That is a catastrophic demand destruction — roughly 3.5 million ounces less demand per year from a market that produces ~7 million ounces. Combined with platinum substitution in the remaining ICE cars, palladium faces a structural surplus. This is why palladium has been declining since 2022 and is the most bearish precious metal. The only offset: potential recycling decline as fewer catalytic converters are scrapped.
| ETF | Metal | AUM | Expense | Structure |
|---|---|---|---|---|
| GLD | Gold | $70B+ | 0.40% | Physical gold in HSBC London vaults |
| IAU | Gold | $30B+ | 0.25% | Physical gold, lower expense than GLD |
| SLV | Silver | $12B+ | 0.50% | Physical silver in JPM London vaults |
| PPLT | Platinum | $1B | 0.60% | Physical platinum |
| PALL | Palladium | $250M | 0.60% | Physical palladium |
Gold mining stocks provide 2-3x leverage to gold prices through operational leverage. If gold rises 10%, a miner's revenue rises 10% but costs (labor, energy, processing) stay relatively fixed, so profits can rise 20-30%. The reverse applies in downturns.
| Ticker | Company | Production | AISC ($/oz) | Leverage to Gold |
|---|---|---|---|---|
| NEM | Newmont Corporation | 5.9M oz/year | ~$1,400 | Moderate (largest, most diversified) |
| GOLD | Barrick Gold | 4.1M oz/year | ~$1,350 | Moderate-High |
| AEM | Agnico Eagle | 3.5M oz/year | ~$1,250 | High (low-cost producer) |
| WPM | Wheaton Precious Metals | Streaming model | N/A (stream) | High (pure financial leverage) |
| FNV | Franco-Nevada | Royalty model | N/A (royalty) | Moderate (diversified royalties) |
| GDX | VanEck Gold Miners ETF | Basket of 50+ miners | Varies | High (diversified, ~2x gold beta) |
| GDXJ | VanEck Junior Gold Miners | Basket of juniors | Varies | Very High (3-4x gold beta, higher risk) |
All-In Sustaining Cost (AISC) is the total cost for a gold miner to produce one ounce of gold, including mining, processing, overhead, and sustaining capital expenditure. It is the miner's effective breakeven price.
At gold $5,125/oz and average AISC of $1,350, gold miners are earning $3,775 per ounce in margin — a record. This margin is why gold miners have been among the best-performing stocks in 2025-2026. The key insight: margin percentage is what matters for stock performance. A $200 increase in gold from $5,000 to $5,200 is only a 4% price increase but a 5.3% increase in margin — operational leverage at work.
Question: Gold rises from $5,000 to $5,500 (10%). A gold miner has an AISC of $1,400/oz. What is the approximate percentage increase in the miner's profit per ounce?
Answer: Profit at $5,000: $5,000 - $1,400 = $3,600/oz. Profit at $5,500: $5,500 - $1,400 = $4,100/oz. Profit increase: ($4,100 - $3,600) / $3,600 = 13.9%. So a 10% gold move produced a ~14% profit increase — 1.4x leverage. With overhead costs and exploration spending, the EPS leverage is typically 1.5-2.5x for major miners and 3-4x for juniors. This is why GDX (majors) tends to move 2x gold and GDXJ (juniors) moves 3-4x gold in bull markets.
The gold/silver ratio measures how many ounces of silver it takes to buy one ounce of gold. It is one of the oldest and most reliable mean-reverting indicators in commodity markets, with a historical range of 40-90 and a long-term average around 65.
| Ratio Level | Signal | Trade | Historical Precedent |
|---|---|---|---|
| > 80 | Silver extremely cheap relative to gold | Long silver / Short gold (or long silver outright) | Ratio hit 120 in Mar 2020, then collapsed to 65 as silver surged |
| 65-80 | Normal range | No clear ratio trade | Most of time spent here |
| < 50 | Silver expensive relative to gold | Long gold / Short silver (or reduce silver) | Ratio hit 32 in Apr 2011 at silver's $50 peak |
A ratio above 80 is historically a strong signal to go long silver relative to gold. The trade can be implemented several ways:
The trade has a strong fundamental basis: when the ratio is above 80, silver's industrial demand provides a floor (unlike gold), and any improvement in gold sentiment will flow disproportionately into silver (smaller market, higher beta).
Question: Gold is at $5,125 and silver is at $58. What is the gold/silver ratio? Is silver cheap or expensive?
Answer: Ratio = $5,125 / $58 = 88.4. This is above 80, putting it in the "silver cheap" zone. Historically, ratios above 80 have been followed by significant silver outperformance as the ratio reverts toward 65. If the ratio compressed to 65 with gold unchanged, silver would be $78.85 (36% upside from $58). If gold continued rising to $5,500 and the ratio compressed to 65, silver would be $84.60 (46% upside). This is a strong setup for the silver catch-up trade.