Global macro · 1929 → today
Every big financial crisis runs in two halves. First a panic, when everyone wants cash. Then a rescue with freshly printed money, when gold recovers first, stocks next and home prices last. This page follows cash, gold, homes and stocks every day since 1927, with every crisis marked on one timeline.
Data as of 2026-11-13
The boom's last stretch. Rates are going up, borrowing gets expensive and home sales slow down first. Stocks are usually still rising.
The panic. People and funds sell almost everything to raise cash, even gold. Cash, especially US dollars, does best.
The day the central bank or government promises unlimited support: rates to zero, printing money, huge spending. That is usually the turning point.
The recovery, paid for with new money. Cash buys less over time. Gold usually rises first, then stocks, and home prices last.
Eleven warning lights, each checked against the levels that showed up near past turning points. Green means calm, amber means keep an eye on it, red means it looked like this in past crises. It is a checklist, not a forecast.
What the warning lights say together
No crisis signal
Markets are calm and the central bank is not in emergency mode. Most of history looks like this.
The S&P 500 compared with its all-time high.
-3.3%
All-time high on 2026-10-06
10% down is a correction, 20% down is a bear market
Data as of 2026-11-13
When even gold is sold, investors are selling whatever they can to raise cash — the peak of a panic.
-3.6% / -3.5%
Worst 20-day stretch in the last 3 months (stocks / gold), ending 2026-11-10
Red when, over 20 trading days, stocks fall 8%+ and gold 5%+
Data as of 2026-11-13
How much big swings investors expect in US stocks over the next month.
19.3
It hit 80.9 in 2008 and 82.7 in 2020
Above 20: nervous. Above 30: panic
Data as of 2026-11-13
The gap between riskier company bonds and US government bonds. It widens when lenders get scared.
1.48%
About 6.0% at the 2008 peak, 4.3% in 2020
Above 2.5%: watch. Above 3.5%: lenders are scared
Data as of 2026-11-12
In a panic the world scrambles for dollars, so the dollar index jumps.
+3.8%
Dollar index at 103.4
Up 3%+: watch. Up 6%+: global dollar shortage
Data as of 2026-11-13
10-year minus 2-year US government bond yields. Below zero ("inverted") means the market expects a slowdown.
+0.27%
Now: normal · lowest in 2 years 0.02%
Inverted = late cycle. Turning positive again after an inversion has often come just before a recession
Data as of 2026-11-13
Whether the Fed is raising or cutting its main interest rate.
3.88%
Peak 5.33%; first cut 2024-09-19; cut 145 basis points (1bp = 0.01%) so far; US stocks since the first cut +32.3%
The first cut is not the bottom. Rates near zero = rescue mode
Data as of 2026-11-12
How fast the Fed's holdings grow. It grows when the Fed creates money to buy bonds (QE).
-0.0%
$6.76tn · flat
Up 1.5%+: adding money. Up 5%+: crisis-scale printing
Data as of 2026-11-11
Case-Shiller national home-price index.
+2.1%
Monthly, published about 2 months late
Below 0: watch. Down 3%+: housing downturn
Data as of 2026-08-01
BIS residential property price index for China.
-6.3%
Quarterly · from peak -21.9%
Below 0: watch. Down 3%+: downturn
Data as of 2026-01-01
The board reads the whole market. PickSkill's AI analyst can apply the same framework to your own holdings and tell you which ones are exposed in this phase.
All the data on one time line, from 1927 to today. Drag to move, pinch to zoom, and point at the chart to read every value on that date. Shaded bands show the phases of each crisis; dots mark key events.
Pick a crisis
Click one and the timeline zooms to it. The number is how far the local stock market fell.
Showing % change since the left edge of the chart, so different assets are easy to compare.
Showing how far each line is below its own highest point so far (0% = at a record).
Showing the actual price. The scale is stretched so a doubling looks the same size whether from 10 to 20 or 1,000 to 2,000.
Showing how far each line is below its own highest point so far (0% = at a record).
Showing % change since the left edge of the chart, so different assets are easy to compare.
What 12 crises have in common, and the order in which cash, gold, stocks and homes take turns doing best.
The boom's last stretch
Panic · everyone wants cash
"Whatever it takes"
Recovery on new money
Prices start rising again
In the panic everything is sold for cash — even gold (−30% in 2008, −12% in March 2020) because it is the most liquid thing to sell for margin. Stocks and gold falling together is the climax of the first half, usually days to months from the low — not the start.
The S&P fell another ~40% after the first cut in 2001 and peaked three weeks after the first cut in 2007. The turns came with QE1 (2008–09), "whatever it takes" (2012), unlimited QE (23 March 2020 — the exact low) and devaluation (1933). Precautionary cuts without a recession (1995, 1998, 2019) are a different animal: stocks rose.
2008: gold Oct 2008, stocks Mar 2009, US housing early 2012. 2015–16: gold Dec 2015, Chinese stocks Jan 2016. Tops run the other way: housing first (2006 US, 1991 Japan), stocks next, gold last — late in the second half when real rates turn up (1980, 2011).
When property or credit is the epicentre (1929, Japan 1990, Hong Kong 1997, 2008, China 2021) housing falls longest — 5 to 15 years — and bottoms last. When it is not (2000, 2015, 2020), lower rates flow into housing and it becomes the second-half winner.
Deflation or inflation? Cash wins debt deflations (1929, Japan, the first half of 2008) and loses inflationary crises (1970s, 2021). Whose cash? In a crisis the reserve currency rises; emerging-market cash is often devalued (1997).
When savings start earning more than inflation (a rising "real" interest rate), holding gold, which pays nothing, looks worse. Gold's big tops came exactly then: Volcker's hikes in 1980 and after "whatever it takes" in 2011–13. In the 2022 hikes record gold buying by central banks was the exception that kept it up.
Twelve crises, each split into the same phases, with how cash, gold, homes and stocks did. Open one for the story, or jump to it on the timeline.
1929–1933
1929–1933
Debt deflation · bank runs · gold standard
Cash BestGold GoodHousing PoorStocks Worst
A margin-fuelled boom ended in the October 1929 crash; the Fed then let a third of US banks fail and, in October 1931, raised rates to defend the gold standard in the middle of a deflation. Prices fell about a quarter, unemployment reached 25%. The spring-1932 open-market purchases briefly stopped the slide; the real turn came when Roosevelt closed the banks (March 1933), left gold and devalued the dollar.
CashBest
CPI −27% → real value of cash +35%
In a debt deflation, cash (in banks that survived) was the best asset of the first half. The catch: one in three banks failed, so "cash" meant physical notes or Treasuries.
GoldGood
$20.67 fixed → $35 (+69%) in 1934
Pegged price, so gold's real value rose with deflation; Homestake Mining stock rose ~6×. Then the state revalued gold by decree — the purest example of "the second half is currency debasement". Private holders had been forced to sell at $20.67 first (EO 6102).
HousingPoor
Nominal ≈ −25–30%, mass foreclosure
Five-year balloon mortgages could not be rolled; foreclosures peaked in 1933. The HOLC (1933) and FHA (1934) created the long amortising mortgage.
StocksWorst
S&P −86% (Sep 1929 → Jun 1932)
Rallies of 20–50% inside the bear market fooled buyers repeatedly. The bottom came with the 1932 Fed purchases; the market then doubled within a year of devaluation.
Key dates
Takeaway
Rate cuts without money creation do not end a debt deflation; printing and devaluation do. Hiking into a deflation (1931) is the textbook policy error.
No house-price, rate or VIX series reach back this far; gold is monthly.
1973–1982
1973–1982
Inflation · end of Bretton Woods · oil embargo
Cash WorstGold BestHousing GoodStocks Poor
Nixon closed the gold window in 1971, the 1973 oil embargo quadrupled oil, and the Fed eased every time unemployment rose. Inflation came in waves (12% in 1974, 14.8% in 1980). This is the cycle that breaks the deflation playbook: the "watershed" was not printing but Volcker's 1979 tightening, which ended the gold bull and started the 1982–2000 equity bull.
CashWorst
Real value of $1 −55% (1971–82)
Deposit rates lagged inflation for most of the decade. Cash only became attractive once Volcker pushed real rates sharply positive (1980–81).
GoldBest
$35 → $850 (Jan 1980), with a −47% mid-cycle drop
Gold rose to ~$195 by end-1974, halved to $103 in Aug 1976 when inflation briefly cooled, then went parabolic. It peaked within weeks of Volcker's shift and then lost ~65% over 20 years.
HousingGood
Nominal roughly ×2.5, real ≈ flat
Housing was the middle class's inflation hedge — especially with fixed-rate mortgages that inflation eroded.
StocksPoor
S&P −48% (1973–74); flat nominal 1968–82, −60% real
P/E compressed to ~7× as rates rose. Equities and gold moved in opposite directions — unlike deflationary crises where both get sold in the panic.
Key dates
Takeaway
Check first whether the crisis is "too little money" or "too much". In an inflationary crisis cash and bonds lose, gold and real assets win, and the turning point is a credible tightening — not a rate cut.
1987
Rising bond yields (10Y above 10%), a falling dollar and "portfolio insurance" selling produced a −20.5% day on 19 October. The Fed under Greenspan announced on the 20th that it stood ready to supply liquidity, banks were told to keep lending, and there was no recession. A pure liquidity panic resolved in days — the prototype of the "Fed put".
CashGood
Best for ~2 months
Cash won the panic but lost the recovery — no deflation followed.
GoldGood
~+5% into December
Gold barely sold off — the crash was an equity-only event, not a system-wide margin call.
HousingMixed
No immediate effect
US housing kept rising until the 1990 S&L-era slump.
StocksWorst
S&P −33.5% (Aug → Dec 1987), new high mid-1989
Without a credit crisis or recession, the crash was a buying opportunity within the year.
Key dates
Takeaway
Not every crash is a crisis. Without credit stress or recession there is no "second half" — a fast liquidity backstop ends it.
1990–2012
1990–2012
Balance-sheet recession · property epicentre
Cash BestGold MixedHousing WorstStocks Worst
At the 1989 peak Tokyo land was said to be worth more than all of the US. The BOJ hiked from 2.5% to 6% (1989–90), the Ministry of Finance capped property lending, and both bubbles burst. Policy then moved too slowly: zero rates only in 1999, QE in 2001, banks recapitalised in 2003. Without a decisive printing moment there was no real second half until Abenomics (2013) — the cleanest counter-example to the usual cycle.
CashBest
Deflation + yen ¥160 → ¥80/$ (1990–95)
With prices flat-to-falling for two decades, yen cash quietly outperformed almost every domestic asset.
GoldMixed
USD gold ~$400 → $255 (1990–2001)
A local crisis in a strong-currency country: gold did nothing for yen holders until the 2000s.
HousingWorst
Commercial land in 6 big cities −87% (1991–2005)
Land peaked in 1991, a year after stocks, and fell for ~15 straight years. Property was the epicentre, so it was first to top and last to bottom.
StocksWorst
Nikkei −80% (1989 → 2003), −82% by 2009
The 1989 high was only regained in February 2024, 34 years later.
Key dates
Takeaway
Rate cuts that come late and small, with zombie banks left in place, turn a crash into a lost generation. When property is the epicentre, it bottoms last — sometimes a decade after stocks.
1997–1998
1997–1998
Currency crisis · dollar debt · contagion
Cash MixedGold MixedHousing WorstStocks Worst
Thailand, Indonesia and Korea had borrowed in dollars against pegged currencies. When the baht was floated on 2 July 1997 the pegs fell like dominoes; Hong Kong defended its peg with overnight rates near 300%, crushing stocks and property. In August 1998 the HKMA bought ~HK$118bn of stocks and Russia defaulted; LTCM's rescue and three Fed cuts ended the global leg.
CashMixed
USD cash king; baht/won/rupiah −50–80%
The cash rule depends on whose cash: local-currency savings were devalued, US dollars were the best asset in the region.
GoldMixed
Up sharply in won/baht; USD gold fell to $252
Korea's 1998 gold-collection drive raised ~227 tonnes from households. In dollars gold kept falling on central-bank sales — gold protects against your own currency, not the dollar.
HousingWorst
Hong Kong −66% (1997 → 2003)
A currency peg means the adjustment comes through deflation of local prices instead of the exchange rate — HK property took six years to bottom.
StocksWorst
HSI −60% (Aug 1997 → Aug 1998); S&P −19% (Jul → Oct 1998)
The HKMA intervention day marked the Hang Seng low; US stocks rallied to new highs within months of the Fed's cuts.
Key dates
Takeaway
In emerging-market crises "cash is king" means dollars. A hard peg pushes the whole adjustment onto local asset prices, so property falls further and longer.
2000–2002
2000–2002
Equity valuation bubble · mild recession
Cash GoodGold BestHousing BestStocks Worst
Nasdaq hit 5,048 on 10 March 2000 after the Fed had hiked to 6.5%. The Fed began cutting with an emergency move on 3 January 2001 — and stocks fell for almost two more years, through 9/11 and the Enron and WorldCom frauds. Rates went to 1% by June 2003. The cheap money flowed into housing, which rose all the way through: the next bubble was born in this crisis's second half.
CashGood
T-bills beat stocks by ~50 points
A cash buffer that captured falling rates was the simplest winning trade of 2000–02.
GoldBest
Bottom ~$256 (Apr 2001) → +40% by 2003
Gold bottomed during the first cuts, years before stocks, and began a ten-year bull market as the dollar topped.
HousingBest
US homes +~20% (2000–02)
Not the epicentre, so housing was a beneficiary of lower rates — the pattern repeats in 2020.
StocksWorst
S&P −49%, Nasdaq −78%
The first cut was nowhere near the low: the S&P fell another ~40% after it.
Key dates
Takeaway
In a recession the first cut is a warning, not a bottom signal. Whatever was not the epicentre (here housing) becomes the next bubble.
2007–2009
2007–2009
Credit crisis · housing epicentre · deleveraging
Cash BestGold GoodHousing WorstStocks Worst
US home prices peaked in mid-2006; subprime losses froze funding markets in August 2007. The Fed's first cut (18 Sep 2007) came three weeks before the stock-market top. After Lehman (15 Sep 2008) everything was sold for dollars — gold fell 30% alongside stocks. QE1 (25 Nov 2008) and its expansion (18 Mar 2009) were the watershed: gold had already bottomed, stocks bottomed on 9 March 2009, and US housing only bottomed in 2012.
CashBest
Dollar index +25% (Jul 2008 → Mar 2009)
The global dollar shortage made USD cash the only thing that went up in the first half. Once QE began, cash rates went to zero for seven years.
GoldGood
$1,011 → $712 (−30%) → $1,921 (Sep 2011)
Sold with stocks in the Lehman margin calls, but bottomed in October 2008 — five months before stocks — and then more than doubled on QE.
HousingWorst
Case-Shiller −27% (2006 → 2012)
The epicentre: first to peak (mid-2006), last to bottom (early 2012), three years after stocks.
StocksWorst
S&P −57% (Oct 2007 → Mar 2009)
The bottom came within days of the QE expansion and the mark-to-market relaxation, as credit spreads peaked.
Key dates
Takeaway
The full sequence in one cycle: housing tops first, first cut near the stock top, stocks and gold crash together at the climax, QE is the watershed, then gold → stocks → housing bottom in that order.
2010–2012
2010–2012
Sovereign debt · bank-sovereign loop
Cash MixedGold GoodHousing PoorStocks Poor
Greece's hidden deficit led to bailouts that spread to Ireland, Portugal, Spain and Italy as weak banks and weak governments dragged each other down. The ECB even hiked in 2011. Gold peaked on 6 September 2011 at the height of the panic; Draghi's "whatever it takes" (26 July 2012) and the promise to buy unlimited government bonds if needed (OMT) ended it without a euro ever being bought. Gold then fell for three years as real rates rose.
CashMixed
Swiss franc +30% until the SNB cap
Cash was only safe in the right currency: German, Swiss and US deposits attracted flight capital from the periphery.
GoldGood
Peak $1,921 (Sep 2011) → $1,050 (Dec 2015)
The textbook late-second-half top: gold peaked at the moment of maximum fear and then lost 45% once the backstop worked and real rates turned up.
HousingPoor
Spain −37%, Ireland −50% from peaks
Where property booms had funded the banks, housing kept falling until 2013–14.
StocksPoor
DAX −33%, S&P −19% (2011)
Stocks bottomed in October 2011 after the LTRO was signalled; the 2012 backstop started a multi-year rally.
Key dates
Takeaway
A credible unlimited backstop can work by announcement alone. Gold's cycle top coincides with peak fear plus the backstop, not with the start of the crisis.
2015–2016
2015–2016
Margin bubble · capital outflow
Cash GoodGold BestHousing BestStocks Worst
Rate cuts from November 2014 and grey-market margin lending sent the Shanghai Composite up 150% in a year. The deleveraging crash from June 2015 was met with a "national team" rescue; the 11 August RMB devaluation then triggered capital flight, and the January 2016 circuit breaker caused a second leg. The turn came from policy: capital controls, the Shanghai Accord and a massive housing destocking push that started a new property boom.
CashGood
RMB −6% vs USD
RMB cash beat stocks easily, but dollar cash beat both.
GoldBest
USD gold bottomed $1,050 (Dec 2015) → +28% in 2016 H1
Gold bottomed one month before Chinese stocks, in line with "gold bottoms first".
HousingBest
Tier-1 cities +30–50% (2015–16)
Housing was the policy tool, not the casualty: easing flowed into property after stocks broke.
StocksWorst
SSE −49% (Jun 2015 → Jan 2016)
Leverage-driven: the fall was as fast as the rise. The 5,178 peak still has not been regained.
Key dates
Takeaway
Leverage-driven equity crashes are fast. Where housing is the policy lever, a stock crisis can make property the next winner.
2020
The fastest bear market on record: −34% in 23 trading days. In the week of 9–19 March even Treasuries and gold were sold as funds scrambled for dollars. The Fed went to zero on 15 March and, on 23 March, announced unlimited QE plus corporate-credit facilities — the same day as the stock-market low. The CARES Act followed four days later. Gold made a new all-time high in August, US housing boomed and inflation arrived in 2021.
CashGood
DXY +8% in 11 days, then −13%
Cash was king for about two weeks; then M2 grew 25% in a year and cash lost purchasing power fast.
GoldBest
−12% in March → record $2,075 (Aug 2020)
The 2008 pattern compressed into weeks: sold in the dash for cash, bottomed first, then led the reflation.
HousingBest
US homes +40% (2020 → 2022)
Not the epicentre, so zero rates turned into a housing boom.
StocksMixed
S&P −34%, new high by August
A five-month round trip: the watershed and the bottom were the same day.
Key dates
Takeaway
When fiscal and monetary bazookas fire together, the watershed comes in weeks, not months — and stocks and gold falling together is the climax, not the beginning.
2022–2023
2022–2023
Tightening crisis · stocks and bonds both down
Cash BestGold GoodHousing MixedStocks Poor
US inflation hit 9.1% and the Fed hiked from 0.25% to 5.5% in 16 months, the fastest since Volcker. Stocks and long bonds fell together, the dollar index hit 114 and the UK gilt market broke in September. The yield curve inverted deeply from July 2022. Silicon Valley Bank's failure (March 2023) brought a liquidity backstop (BTFP), not cuts; gold, with record central-bank buying, broke out to new highs.
CashBest
T-bills ~5%; DXY +18% to Sep 2022
In a tightening crisis cash is the winner because its yield rises — the opposite of the 1970s, when the Fed stayed behind inflation.
GoldGood
$2,070 → $1,620 (−21%) → new highs from 2024
Rising real rates usually crush gold; this time central-bank buying (after Russia's reserves were frozen) limited the fall.
HousingMixed
US −4% then new highs; China −10–20% (2021+)
US owners with 30-year fixed mortgages did not sell; turnover collapsed instead of prices.
StocksPoor
S&P −25%, Nasdaq −35%
Valuation compression without credit stress; stocks bottomed when inflation and yields peaked in October 2022.
Key dates
Takeaway
A tightening crisis has its own playbook: cash wins, bonds lose, and the turn comes when inflation peaks. The usual deflation playbook (buy gold at the first cut) does not apply.
2021–
2021–
Property deleveraging · local debt · deflation pressure
Cash GoodGold BestHousing WorstStocks Poor
The August 2020 "three red lines" cut off developers' credit; Evergrande defaulted in late 2021 and sales and prices fell for years. It looks most like Japan after 1990 plus the housing leg of 2008: property is the epicentre, inflation slid towards zero and rate cuts were gradual. The 24 September 2024 package (rate cuts plus facilities to fund stock buying) was the first "whatever it takes"-style signal for equities; housing has not yet confirmed a bottom.
CashGood
CPI ≈ 0%; deposits beat property
Japan-style: with near-zero inflation and falling home prices, household savings piled into deposits.
GoldBest
RMB gold more than doubled (2022 → 2025)
PBOC buying from late 2022 and household demand made gold the best Chinese asset of the cycle.
HousingWorst
Resale prices −30%+ in many cities
Epicentre: first to fall, slowest to recover — watch whether a large central-government balance-sheet backstop arrives.
StocksPoor
HSI −55% (2021 → 2022), SSE −27% (2021 → 2024)
Stocks turned on the September 2024 package, ahead of housing — consistent with "stocks bottom before property".
Key dates
Takeaway
An ongoing case. The model says watch for the second half's signature: large-scale central balance-sheet expansion, not another rate cut.
Phase dates for an ongoing cycle are provisional.
Put the pattern to work
This page is free. PickSkill turns the same data into decisions about what you actually own.
Ask in plain language: "Is this like 2008?", "How much gold should I hold now?" Answers combine market data, filings and news, with sources.
Ask a questionAdd your holdings across US, Hong Kong and A-shares and see trend, momentum and money-flow signals for each one every day.
Build a portfolioTurn any analysis into a PowerPoint or Word research report you can share, in minutes.
Create a reportConnect the PickSkill MCP server and ask for signals, screens and your portfolio from the AI tools you already use.
Set up MCPFree plan: $0, 30 AI conversations and 5 deep-research runs a month, 1 portfolio. Paid plans from $15/month.
Sources
Only if there is no recession. In recessionary cycles (2001, 2007) the S&P 500 fell a further 40–50% after the first cut. The more reliable turning point has been the moment policy goes unlimited — QE, zero rates plus a fiscal package, or a devaluation.
Because it is liquid and usually in profit. In a margin-call phase investors sell what they can, so gold fell about 30% in 2008 and 12% in March 2020. In both cases it bottomed before stocks and then made new highs once central banks started printing.
In the first half of a deflationary crisis, yes — especially US dollars. In the second half, or in an inflationary crisis, cash loses purchasing power. Holding cash in a currency that is being devalued (Asia 1997) is not safe.