Where the Money Is.A 57-pair ratio map of US equity, credit, FX, commodity, and international flows.

Each of the eleven SPDR Select Sector ETFs (each one holds the largest companies in its sector, weighted by market value) is measured as a ratio against SPY. SPY is the SPDR S&P 500 ETF, the broadest large-cap US benchmark: every dollar that goes into SPY is implicitly bought across all 500 names, weighted by market value. SPY itself returned +26.1% over the window, so the bar for any sector to be "outperforming" is steep. A rising ratio means the sector is winning capital relative to the market. A falling ratio means it is losing capital relative to the market. Range position tells you where today's value sits inside the year's high-to-low band (0% means at the year's low, 100% means at the year's high).

Only two of eleven sectors cleared SPY's bar. XLK (Technology) at +19.7% is the lone aggressive winner. Tech's absolute return of +50.9% nearly doubled the index. XLE (Energy) is a distant second at +6.8%. Everything else lost ground vs SPY, which means the rally was extraordinarily narrow and was carried almost entirely by mega-cap tech.

Classic defensives capitulated. Capitulation here means the last holders give up and sell into the weakness, which usually marks a near-term low because nobody is left to sell. XLU (utilities, the bond-proxy sector whose share prices behave like long-duration Treasuries) is at -12.5%. XLP (consumer staples) is at -17.2%. XLRE (real estate, holding rate-sensitive Real Estate Investment Trusts whose dividends fall in relative value when bond yields rise) is at -15.5%. Investors did not want bond-proxies in a risk-on tape. Worst of all is XLF (financials) at -21.1%. Banks could not keep up despite a strong market.

Direction matters. XLK is still accelerating: its 1-month change (+11.8%) is running faster than its 1-year pace, which means the tech bid is intensifying rather than fading. Conversely XLE has flipped negative on 3-month and 1-month despite its positive 1-year. Energy leadership is rolling over right now, a rotation-in-progress signal.

Range positioning tells. Range position is where today's ratio sits inside the 1-year high-to-low band. Above 90% means near the high (still strengthening). Below 10% means near the low (capitulation, possible inflection point if the trend reverses). XLK at 94.7% is pinned to its 1-year high with positive 1-month direction, so no top is visible yet. XLU, XLY, and XLF all sit at literal 0.0% range position, meaning their ratios versus SPY are at 1-year lows with 1-month still negative. That is capitulation. The inflection has not yet occurred.

Regime readHyper-narrow tech-led melt-up with broad defensive and financial capitulation. Late-cycle momentum regime, not yet rotating.

Six diagnostic ratio pairs. XLY/XLP (cyclical sectors vs defensive sectors) is the headline appetite gauge. Discretionary spending (XLY, restaurants, cars, hotels, online retail) outperforms consumer staples (XLP, toothpaste, soap, groceries) when households feel confident enough to spend on wants rather than just needs. XLK/XLU contrasts tech with utilities. Utilities are a bond-proxy sector, meaning their share prices behave like long-dated Treasury bonds: regulated utility revenue and dividends are stable and predictable, so investors price them off the prevailing 10-year and 30-year bond yield. When yields fall, both bonds and utility shares rally. When yields rise, both fall. IWM/SPY is small-caps vs large. Small caps lead in genuine risk-on tape because investors will only own riskier, less-liquid names when they are confident. IWF/IWD is growth-style stocks vs value-style stocks. RSP/SPY is the market-breadth gauge. RSP equal-weights every S&P 500 name (every company is held in equal proportion regardless of size), so when it lags cap-weighted SPY (where bigger companies count more), the index gains are concentrated in a few mega-cap names doing all the work. MTUM/USMV contrasts momentum-chasing (buy whatever is going up) with low-volatility defensiveness (buy whatever has been quietest).

The surface tape screams risk-on. XLK/XLU (tech vs the utility bond-proxy) is pinned at the 1-year high, at 100% range position, with +36.8% on the year. MTUM/USMV (momentum-chasing names vs low-volatility names) sits at the 97.5th percentile of its 1-year band with +31.5%. Both are at extremes.

But the cleaner gauges disagree. XLY/XLP, the headline cyclicals-vs-defensives signal, is only +5.5% on the year, sitting mid-range (36.7% of its band), and crucially has rolled over to -2.0% in the last month after a +5.2% three-month run. The consumer-discretionary trade is rolling over. That is the classic late-cycle topping signature: investors stop paying up for cars, hotels, and online retail and start defending cash flows in toothpaste and groceries.

Breadth is awful. Market breadth is the share of stocks participating in a rally; healthy breadth means most names go up together. RSP/SPY sits at the absolute 1-year low (range position 0.0%) and is still bleeding -4.9% on the month. RSP equal-weights all 500 S&P names (every company counts the same), so when it lags cap-weighted SPY (where bigger companies count more) this hard, the index gains are concentrated in a narrow cohort of mega-cap names. The entire risk-on signal is being carried by that cohort.

IWF/IWD (large-cap growth-style stocks vs large-cap value-style stocks) is dead flat on the year (+0.1%) but quietly turning up over the last 3 months (+4.0%). That is a late-stage mega-cap-growth squeeze, not a durable trend reversal. IWM/SPY (small caps vs the S&P 500) rallied for most of the year (+8.0%, range pos 76.5%) but has stalled and reversed (-1.6% on the month).

Regime readRisk-on by index optics, but it is a narrow mega-cap-tech-and-momentum trade. Breadth, small caps, and the XLY/XLP cyclical gauge are all turning defensive underneath.

These ratios reveal what is leading inside each sector. Even sectors that look weak in Group A can have hot internal rotations. SMH and IGV inside Tech (semiconductors vs software); XBI, IHI, IHF inside Healthcare (biotech, medical devices, healthcare service providers); KRE, KBE, KIE, IAI inside Financials (regional banks, all-banks, insurance, broker-dealers and exchanges); ITB and XRT inside Discretionary; JETS and IYT inside Industrials; XOP and OIH inside Energy; GDX inside Materials. The rotations within sectors often run earlier and harder than the parent sector itself.

Tech. Complete divergence inside the sector. SMH/XLK (semiconductors vs broad tech) is ripping +53.5% on the year, at 99.8% of its 1-year range, still accelerating +5.9% on the month. IGV/XLK (software vs tech) is at 0% of range, -43.7% on the year, still falling. Software's absolute return is -15.1% while XLK as a whole is +50.9%. The entire XLK return is silicon, not Software-as-a-Service. The gap is widening, not narrowing.

Healthcare. Three-way story but two of the three are surging. XBI/XLV (biotech vs healthcare) at +55.7%, range pos 94% (XBI absolute return +75%) is classic risk-on small-cap healthcare leadership (XBI equal-weights its biotech holdings so it skews to smaller, riskier names). IHF/XLV (healthcare service providers, mostly managed-care insurers like UnitedHealth and Elevance, vs healthcare overall) is the freshest mover: flat on 1 year but +17.4% in just the last month, at 86% of range. Sharp rotation into managed care. IHI/XLV (medical devices) is the loser, -30.5% on the year, at 0% of range, still bleeding. The rotation is out of devices and into biotech plus providers.

Financials. IAI/XLF (broker-dealers and exchanges, the firms that earn fees from trading, IPOs, and mergers, vs financials overall) is the standout. +16.2% at 99.9% of range, still climbing +3.2% on the month. Capital-markets activity (initial public offerings, mergers and acquisitions, trading-desk volume) is white-hot. KRE/XLF (regional banks) is +14.0% and KBE/XLF (all banks, regional plus mega-bank) is +12.8%. KRE slightly leads KBE, which means regional banks are participating in the bank rally rather than just the mega-banks carrying the index. KIE/XLF (insurance) is the laggard at -5.1%. Note: XLF itself is the worst-performing sector vs SPY in Group A, yet its sub-sectors are nearly all positive. The headline financials weakness is real but the internal picture is constructive.

Discretionary. Both sub-ETFs underperform XLY. ITB/XLY (homebuilders) is at 0% of range, -13.6% on the year, -21.8% on 3 months, -8.7% on 1 month. Housing is in active breakdown. XRT/XLY (retail) at 2.7% of range, -4.1% on the year. XLY's modest absolute gains are coming from mega-cap names (Amazon and Tesla dominate the ETF), not the cyclical sub-industries.

Industrials. Both subs trail XLI (which is +22.1% absolute). IYT/XLI (transportation, the rails and truckers and parcel carriers that move goods) is -4.1% on the year, stabilizing on 1 month. JETS/XLI (passenger airlines) is the weaker leg, -11.9% on 3 months, at 12% of range. Travel demand is fading versus the broader industrials story (defence, aerospace, capital-equipment spending).

Energy. OIH/XLE measures oil services (the firms that drill the wells, manufacture the equipment, and provide the technology to oil companies, the picks-and-shovels of the upstream oil business) versus energy overall. +41.9% on the year, at 94% of range, still climbing +5.4% on the month. XOP/XLE (oil-and-gas exploration and production, the upstream companies that own the wells and produce the oil) is roughly in line with energy as a whole: -0.4% on 1 year but recently turning to +7.7% on 3 months. The capital-spending (capex) theme is decisively beating pure production.

Materials. GDX/XLB (gold miners inside the broader materials sector) is +71.3% on the year. GDX's absolute return is +107% (operating leverage to gold: the gold price moved +44% over the year, but because mining costs are fixed, almost all of the price move falls to miner profit). But the ratio is at 66% of range and cooling (-6.5% on 3 months, -4.6% on 1 month). Gold-miner dominance was the materials trade, but the flight-to-safety bid is now easing.

Internal rotation readRisk-on at the extremes: semiconductors, biotech, capital markets, oil services, and (newly) healthcare providers are leading. Software, medical devices, homebuilders, airlines, and gold miners are the rotation-out trades.

Where the long-running theme money is flowing. Three buckets. AI and Innovation: ARKK/SPY, BOTZ/SMH, HACK/IGV (Cathie Wood's disruptive-innovation fund vs the index, AI and robotics ETF vs raw semiconductors, cybersecurity vs broad software). Energy transition: TAN (solar), ICLN (clean energy), URA (uranium), each measured against XLE (traditional fossil fuel energy). Critical materials and China: LIT (lithium and battery materials) vs XLB (broad materials), KWEB (China internet) vs EEM (broad emerging markets).

AI and Innovation complex. The winner is cybersecurity: HACK/IGV is +24.7% on the year at 88.8% of range. HACK's absolute return is only +5.8%, but IGV (the broad software ETF) is at -15.1%, so this is a "cyber holds, software cracks" story. The +4.0% on the month says it is still rotating in HACK's favour right now.

ARKK/SPY measures Cathie Wood's flagship Disruptive Innovation fund against SPY. ARKK is what is called a high-beta fund, concentrating in unprofitable growth names. Beta here is just how much a fund moves when the market moves: a beta of 1 means it moves in step with SPY, a beta of 2 means it moves twice as much. ARKK has a beta of 2.26 versus SPY, so when SPY moves 1%, ARKK tends to move 2.26%. The ratio is +9.1% on the year but -2.4% in the last month. ARKK's absolute +37.6% is real, but it is now losing to SPY on the recent horizon.

The big loser is BOTZ/SMH at -42.1%, 1.8% of range, -10.2% on the month. BOTZ (Global X Robotics and AI ETF, holding mostly industrial-automation companies) gained +34% absolute, but SMH (the pure semiconductor basket) ripped +131.6%. The AI trade is the chips themselves, not the thematic ETF. If you wanted exposure to AI through a packaged robotics-and-AI fund, you lost. If you owned the chips directly, you doubled.

Energy transition. All three transition themes beat XLE on the year, but they diverge sharply now. TAN/XLE (solar manufacturers and developers vs traditional energy) is the live leader at +32.9% on 1 year, +12.2% on 1 month, range 85.4%. TAN is heavily rate-sensitive because solar projects are debt-financed: the chain is bond yields fall → the cost of financing solar projects falls → solar developer economics improve → TAN rallies before any change in earnings actually shows up. So the late-cycle yield drop is doing the work. ICLN/XLE (broad clean energy) follows at +26.3% on 1 year, +7.7% on 1 month. URA/XLE leads on 1-year absolute (+41.4%, URA absolute +90.3%, the uranium and nuclear-renaissance trade) but is the only one rolling over: -5.3% on 3 months, -2.1% on 1 month. The uranium trade is digesting. Solar is taking the baton.

Critical materials and China. LIT/XLB is the trade of the year: +91.4% on 1 year, 96.5% of range, still +11.1% on the month. LIT (lithium miners plus battery-materials companies) doubled (+131% absolute) while XLB (broad materials) did +20.8%. Electric-vehicle and battery-supply scarcity is back as a market theme. KWEB/EEM (China internet, Alibaba, Tencent, Meituan, PDD, vs broad emerging markets) is at -39.1%, 5.5% of range. KWEB lost 10.8% absolute while EEM gained +46.6%. China internet is being abandoned inside emerging markets.

Thematic readCapital is flowing into the hard-asset plumbing of the energy transition (lithium, solar, clean energy) and into cybersecurity. AI and robotics baskets, ARKK-style innovation, and China internet are all being left for dead in favour of the underlying chips, the S&P itself, and broad emerging markets ex-China.

HYG/LQD is the credit-spread proxy. HYG holds junk bonds: corporate debt rated below investment-grade, paying higher coupons to compensate investors for taking on default risk. LQD holds investment-grade corporates: safer, lower-yielding, mostly the big blue-chip companies. When HYG outperforms, investors are bidding junk over investment-grade, so credit spreads are tightening and the market is risk-on. When LQD outperforms, the bid is flight-to-quality, spreads are widening, and that often runs a few weeks ahead of equity stress. TLT/IEF compares 20-year-plus Treasury bonds (TLT) to 7-to-10 year Treasury bonds (IEF). TLT rising faster than IEF is called bull-flattening (the long end of the curve rallies harder than the middle, often on growth fear or rate-cut anticipation). TLT falling harder than IEF is bear-steepening (the long end sells off harder, usually because investors demand more compensation for tying their money up that long). HYG/TLT and LQD/TLT compare credit (taking corporate default risk) with duration (taking interest-rate risk on Treasuries). When credit outperforms duration, investors prefer yield from credit over yield from rates.

Credit appetite. HYG/LQD is the credit-spread proxy. A credit spread is the extra yield that risky corporate bonds pay over safe US Treasuries. When spreads tighten (junk yields fall toward Treasury yields), the market is risk-on. When spreads widen (junk yields rise above Treasuries), the market is risk-off. HYG holds junk bonds (corporate debt rated below investment grade, BB and lower, paying higher coupons to compensate investors for the higher chance the issuer defaults). LQD holds investment-grade corporates (safer than HYG, riskier than Treasuries). The ratio is -0.8% on the year (a modest credit-spread widening, which often runs ahead of equity stress by a few weeks), but +0.8% on 3 months and +0.5% on 1 month. The short-term direction is contradicting the 1-year read. Spreads have been tightening recently. Credit-risk-on regime has returned over the past quarter, even though it has not yet erased the full-year wound.

Curve shape. TLT/IEF sits at 6.2% of its 1-year range, essentially pinned at the lows. TLT is the iShares 20+ Year Treasury Bond ETF (long-duration government bonds). IEF is the iShares 7-10 Year Treasury Bond ETF (intermediate-duration government bonds). Duration here means how much a bond's price moves when interest rates change: longer-maturity bonds have higher duration so they swing more for any given yield move. The 10-year has outperformed the 20-year across every window (1 year, 3 months, 1 month all negative on TLT/IEF). This is bear-steepening: the long end of the yield curve sells off harder than the middle, which steepens the curve, but for a worrying reason. The chain is investors demand more compensation (term premium) for tying their money up for 20 or 30 years → long bonds sell off → long yields rise faster than middle yields → the curve steepens. The motivation is usually inflation worries, large Treasury issuance to fund deficits, or generic uncertainty about the future. If the opposite were happening (TLT leading IEF), the curve would be bull-flattening: long bonds rallying harder on growth fear or expected rate cuts.

Cross-asset implications. Both HYG/TLT and LQD/TLT sit at 100% of their 1-year range. Both forms of credit (junk and investment-grade) are beating long-duration Treasuries. Investors want yield from corporate-default risk (credit), not yield from interest-rate risk on government bonds (duration). The classic recession hedge of owning long Treasuries is being abandoned.

Credit readThe bond market is confirming risk-on equity flows on the short-term horizon (credit spreads tightening, duration shunned), but the year-long bear-steepening of the Treasury curve is a quiet warning. The marginal buyer of long bonds is demanding more term premium, which means inflation worries or fiscal worries are running underneath the risk-on tape.

SPY/TLT is the cleanest stocks-vs-bonds read. Rising means equity bid. Falling means bond bid, usually because investors are pricing in either a recession or imminent Fed rate cuts. GLD/SPY rising means gold is being hedged even while stocks rally, a debasement signal (debasement here means the purchasing power of the dollar is eroding, so people want a hard asset instead of paper). GDX/GLD measures gold-miner operating leverage. Operating leverage just means that when most of a company's costs are fixed, a small revenue move flows almost entirely to profit. A gold miner's revenue is the gold price; the mining cost (digging the dirt, refining the ore, paying labour) is roughly fixed. So a 44% gold-price move can deliver a 100%-plus move in miner earnings and miner equity. If GDX outpaces GLD, the gold trade has equity-style speculative conviction. If miners lag gold, the bid is purely defensive store-of-value. GLD/SLV is the classic gold-silver ratio. Silver is gold's industrial cousin: roughly half of silver demand is solar panels and electronics, while gold demand is mostly jewellery and central bank reserves. When fear dominates, gold leads silver and the ratio rises. When industrial demand or reflation dominates, silver leads gold and the ratio falls.

Underlying absolute moves over the year: GLD +44.5%, SPY +26.1%, SLV +168%, GDX +107%, TLT -1.1%. Everything risk-on and hedge-on rallied. Only long Treasuries got left behind.

Tickers in plain English. GLD is SPDR Gold Shares, which actually owns physical gold bullion in a vault. SLV is iShares Silver Trust, which actually owns physical silver. Silver is gold's industrial cousin: roughly half of silver demand is solar panels and electronics, while gold demand is mostly jewellery and central bank reserves. GDX is VanEck Gold Miners ETF, which holds the equity (the shares) of gold-producing companies. TLT is the 20+ year Treasury ETF, the long-duration government-bond safe haven.

SPY/TLT at the 1-year high (range 100%), +27.5% over the year and still accelerating (+11.4% on 3 months, +9.9% on 1 month). Equities are crushing bonds. Capital is fleeing duration, not seeking it. This is not a recession-fear tape.

GLD/SPY is +14.6% on the year (gold beat stocks), but the last quarter reversed sharply (-14.2% on 3 months, -9.5% on 1 month, range 44%). The hedge bid was real in the first half of the window but has been unwinding into the equity melt-up. Gold is not being bought as a panic hedge right now. It is being sold down relative to risk.

GDX/GLD is +43.2% at 81% of range. GDX +107% vs GLD +44.5%, so miners delivered roughly 2.4 times gold's move. That is operating leverage in action: a miner's profit margin equals the gold price minus the (roughly fixed) cost of pulling the metal out of the ground, so a 44% gold-price move flows almost entirely to profit. The fact that GDX outpaced GLD by this much means the gold trade has equity-style speculative conviction attached to it, not pure defensive store-of-value buying.

GLD/SLV is -46.1% on the year, 14% of range, still falling hard (-12.2% on the month). Silver (+168%) ran 3.8 times gold over the year. This is the textbook pro-cyclical / debasement / inflation signature: when fear dominates, gold leads silver and the ratio rises. When reflation or industrial demand dominates, silver leads gold (because of its electronics and solar exposure) and the ratio collapses. It just collapsed.

Safety readHedge flows are pro-cyclical debasement, not defensive fear. Metals are rallying as a reflation and anti-fiat trade in lockstep with risk assets, while the classic safe haven (long bonds) is being liquidated.

UUP is the Invesco DB US Dollar Bullish ETF. It tracks DXY, the US Dollar Index, which measures the US dollar's value against a basket of six major currencies (euro, Japanese yen, British pound, Canadian dollar, Swedish krona, Swiss franc), weighted by trade importance. UUP/GLD measures the dollar's real purchasing power against gold. Falling means debasement: the dollar buys less gold over time. UUP/SPY reads the dollar as a haven. It rises in equity drawdowns when the world flees to dollar cash; it falls when investors instead borrow cheap dollars to buy risk assets (in that mode the dollar is the funding currency, not the haven). COPX/GDX is copper miners vs gold miners. Copper is nicknamed "Dr Copper" because copper demand tracks industrial activity in real time. Copper goes into wiring (housing), motors (electric vehicles), grid infrastructure, and AI data centres. When copper miners outperform gold miners, the industrial cycle is expanding.

UUP/GLD collapsed -31.5% on the year. UUP was -1% while GLD was +44.5%. That is a textbook debasement signal: the dollar is losing real purchasing power as global capital flees into the monetary hedge (gold). But the 3-month has flipped sharply positive (+11.4%) and the 1-month is still +4.1%, with range position at 25%. Gold's parabolic move is cooling and the dollar is stabilising off its lows. The structural trend is still down. A tactical bounce is on.

UUP/SPY is at 0.6% of its 1-year range, essentially pinned to the 1-year low. -21.5% on the year and crucially still falling (-4.4% on 3 months, -5.8% on 1 month). No flip. The dollar is being actively sold to fund risk-asset purchases. That is classic late-cycle risk-on flow: investors borrow cheap dollars to buy stocks, gold, and foreign assets, which weakens the dollar further. There is no flight-to-safety bid anywhere. The dollar is the funding leg of the melt-up, not a haven.

COPX/GDX is at 100% of 1-year range, +10.6% on 1 year, still accelerating (+12.3% on 3 months, +11.3% on 1 month). Copper miners +129% versus gold miners +107% on the year: both ripping, but Dr Copper (the market's nickname for copper, because copper demand tracks real industrial activity) is now leading. That is a pro-cyclical, growth-on confirmation.

Macro readWeak dollar plus copper leading gold equals a global cycle in pro-cyclical growth-on mode. The dollar is acting as the funding leg for risk assets, not as a haven.

EEM/SPY measures broad emerging markets against the US. EEM is the iShares MSCI Emerging Markets ETF, with heavy weights in China, India, Taiwan, Korea, and Brazil. EFA/SPY measures developed markets outside the US. EFA is iShares MSCI EAFE (Europe, Australasia, Far East), holding the big developed-market companies that are not US-listed. FXI/SPY isolates Chinese large-caps specifically, which lets us tell whether broad EM strength is China-driven or non-China-driven. EWJ/SPY isolates Japan. EWJ is dollar-denominated but the underlying companies report in yen, so its performance depends on the USD/JPY exchange rate as well as Japanese earnings. A rising ratio means the region is winning capital relative to the US.

EEM/SPY is the only region decisively beating SPY. EEM (broad emerging markets, weighted heavily toward China, India, Taiwan, Korea, and Brazil) returned +46.6% vs SPY's +26.1%, so the ratio is +16.3% on 1 year, parked near its 1-year high (89.9% of range) and still grinding up (+1.0% on 1 month, +1.8% on 3 months). That is a classic global-cycle-on, softer-dollar tape. But as the next paragraph shows, this is not a China story.

EFA/SPY failed the hurdle. EFA (developed-market equities outside the US: Western Europe, Australia, Japan, Hong Kong, etc.) returned +21.4% vs SPY's +26.1%, so the ratio is -3.8% on 1 year. Worse, it has rolled hard: -7.8% on 3 months, -6.3% on 1 month, sitting near 1-year lows (10% of range). Europe and developed Asia were a 2025 darling. That trade has been unwinding sharply.

FXI/SPY is the year's worst international loser. FXI (China large-cap) returned +8.5% vs SPY's +26.1%, so the ratio is -13.9%. The ratio is pinned near its 1-year low at 8.8% of range. The regression of FXI's daily returns on SPY's daily returns produces a negative alpha of -0.051. Alpha here is the constant term from that regression: it is what FXI returns when SPY returns zero. A negative alpha means FXI structurally underperforms SPY even after adjusting for how much it moves with SPY (its beta). Investors are not coming back to China.

EWJ/SPY barely cleared SPY for the full year (EWJ +29.8% vs SPY +26.1%, ratio +2.9%) but has given it back: -7.7% on 3 months, -2.6% on 1 month. The reflation and corporate-reform thesis that drove Japan worked into early 2026 and is now rolling over, likely yen-driven (a weakening yen lifts Japanese exporter earnings in local terms but hurts the dollar-denominated ETF).

The EEM win paired with the FXI loss is the tell. This is the textbook "EM-ex-China" trade: India, Taiwan, Korea, and Brazil are carrying EEM while its single largest country weight (China) bleeds. EFA and EWJ have both flipped negative on 3 months and 1 month despite better 1-year prints, which is a fresh rotation back into US large-cap.

International readThe US is still receiving the marginal global dollar versus Europe, Japan, and China. But it is losing it to emerging Asia ex-China, where EEM has decisively broken out and is the only international bet currently working.