Macro ReviewHow to read

At the the is 42% below the price; no target

A two-stage payout model on this site's 5.8% near-term earnings growth (consensus 15.1%) and a 5.2-point premium values the index at 4,338: a model reading, not a forecast.Where are markets, and what state are they in?

Markets to 16 Nov 2026Every reading datedUpdated 18 Nov 2026
Gap to the model's value−41.5%the model's value at the yield-gap premium 4,338 (−42%) · · no target
S&P 500, year to date+9.4%total return to 16 Nov 2026
10-year Treasury yield5.12%16 Nov 2026 · 106bp above the 4.07% would justify
39.4×Nov 2026 · 98th 1881
Crude oil (WTI), year to date+102.5%oil fund total return to 16 Nov 2026
3of 6 on Risk sentiment · week of 13 Nov 2026

In brief

Where markets stand, and what would change the reading?
  • At the the is 42% below the S&P 500's price (16 Nov 2026).→ Exhibit 1
  • The 10-year sits 106bp above the value would justify (16 Nov 2026).→ Exhibit 5
  • The S&P 500 is +9.4% this year and crude oil +102.5% (to 16 Nov 2026).→ Exhibit 3
  • The regime on the four-regime map is “Only the dollar hedges”, for 37 weeks since 6 Mar 2026.→ Exhibit 5
  • The is 39.4× (98th 1881), to Nov 2026.→ Exhibit 5
What would change this view
What would change this viewTriggerTodayDecided by
The gap to the model's value closesgap to the model's value ≥ 0%−41.5% (16 Nov 2026)not crossedeach valuation update
The 10-year completes the move to a Fed on hold10-year Treasury yield ≤ 4.07%5.12% (16 Nov 2026)not crossedeach Treasury close
The S&P 500 gives up the yearS&P 500 year to date < 0%+9.4% (16 Nov 2026)not crossedeach close
The regime changesanother centre nearer by 0.2 of a typical distance0.63 further away (bonds hedge, calm; week to 13 Nov 2026)not crossedthe week to 20 Nov 2026
The reasoning behind each point
  • The model's value
    At the yield-gap premium the model's value is 42% below the S&P 500's price (16 Nov 2026).

    The model's value at the yield-gap premium 4,338 (−42%) · record unproven · no target. The model is the two-stage fair-value model on the Valuation page — near-term growth from the Earnings page fading to the long-run rate, discounted at the 10-year plus the equity risk premium. A negative gap means the index sits above the model's value.

  • Priced
    The 10-year sits 106bp above the value a Fed on hold would justify (16 Nov 2026).

    The 10-year is 5.12% against the 4.065% fair value of the “Hold at IORB” path at today's term premium, on the What is priced page.

  • Markets
    The S&P 500 is +9.4% this year and crude oil +102.5% (to 16 Nov 2026).

    Both are total returns on the funds the Performance page tracks, each dated because the section's pages do not all end on the same day.

  • Regime
    The regime on the four-regime map is “Only the dollar hedges”, for 37 weeks since 6 Mar 2026.

    When stocks fall, the dollar rises but Treasuries do not, week to 13 Nov 2026. On the daily sign rule, stocks have moved with bonds and against oil for 178 sessions since 5 Mar 2026, so bonds are not hedging equities. To 16 Nov 2026, the 63-day stock–bond correlation is +0.10 and stock–oil −0.36; that sign pattern is read from the signs of the two correlations and named by its legs. The hidden-state model (weeks sorted into 6 recurring co-movement patterns) names the closest pattern “bond hedge weak, oil with stocks”, since 23 Oct 2026, week to 13 Nov 2026; today's stock–bond and stock–oil readings sit outside that pattern's own range, so the name lags. 6 states is the number the Bayesian information criterion picks.

  • Valuation
    The cyclically adjusted P/E is 39.4× (98th percentile since 1881), to Nov 2026.

    Its 20-year mean is 27.9×. From the Valuation page.

  • Volatility
    S&P 500 20-day volatility is 23.2%, the 89th percentile since 2010 (16 Nov 2026).

    On the Positioning page 17 futures positions sit outside their 20th–80th percentiles (crowded positions of 40 · 10 Nov 2026). Realised volatility from the Volatility page; the crowding count from the Positioning page's CFTC reports, which are dated to the Tuesday they describe.

Next: Valuation — what the model's value rests on

Source: Yahoo Finance; U.S. Treasury; Macro Review calculations (the pages linked)cross.overview-read

Markets at a glance

What moved across about two dozen benchmarks, over a day, a week, a month, the year and twelve months

Exhibit 1 · S&P 500 model value on each equity risk premium, against its price

Two-stage model, index points, 16 Nov 2026; dot the central value, bar near-term growth's 10th–90th percentile; vertical line the S&P 500 at 7,417
About this exhibit
The two-stage model's value of the S&P 500 in index points on each equity-risk-premium basis (today's implied premium, the yield-gap premium, the full-history median, the yield-gap premium in months when stocks and bonds moved apart, all inputs drawn). Each bar spans near-term growth's range with the central value as a dot, and a vertical line marks the S&P 500's price.
The model's value by premium basis — the Valuation page's answer
The model's value in index points on each premium basis, 16 Nov 2026; each bar spans near-term growth's 10th to 90th percentile, the dot and the figure are the centre. The one number the site prints is the model's value at the yield-gap premium 4,338 (−42%) · record unproven · no target. No target. Model-conditional.
The Valuation page's football field, copied row by row and compared with it on every update: the same two-stage model and inputs on every row, only the equity risk premium changes. The register's row is the yield-gap premium (today's implied premium plus the yield gap's distance to its median in months when stocks and bonds moved together); the others show how much the answer depends on that choice. The last row draws near-term growth and the premium together — the spread of the inputs, not a probability.
Source: U.S. Treasury; Robert J. Shiller data via multpl.com; FactSet; Federal Reserve; Macro Review calculations (the Valuation page) · data to 16 Nov 2026cross.overview-9

Exhibit 2 · S&P 500's gap to its model value, month by month

Gap to the model's value, %, monthly, Nov 2006 to Nov 2026; below zero the index sits above it; shaded Jun 2007 – Dec 2010, earnings write-downs
About this exhibit
The S&P 500's percentage gap to the two-stage model's value, computed on the data available at each date, monthly. Below zero the index sits above the model's value. The years of the earnings write-downs are shaded.
The S&P 500's gap to the model's value, twenty years
Two-stage fair-value model on the data available at each date (the valuation history's rules), monthly, Nov 2006 to Nov 2026; below zero the index sits above the model's value. Its last reading −52.1%; on the Valuation page's rules −41.5%, the gap the tile prints. Shaded, Jun 2007 – Dec 2010: the 2008–09 write-downs cut trailing reported earnings by 92%, then months the model cannot price (no line). The highest reading after them, +199% in Jul 2012, is the earnings model expecting earnings to rebound from that base — not a valuation signal.
Source: Robert Shiller; U.S. Treasury; Macro Review calculations (the Valuation history page) · data to Nov 2026cross.overview-8

Exhibit 3 · Return or yield change of each benchmark, by window

Total return, %; yields in bp, style spreads in log points; 1 day to 1 year; data to 16 Nov 2026 unless a row is dated
About
AssetLast1d1w1mYTD1yFrom 52w high
US equity
S&P 500SPY$739.96−1.9%−1.7%−3.1%+9.4%+11.3%−5.2
S&P 500 equal weightRSP$218.93−1.0%−0.9%+2.3%+15.7%+18.7%−2.0
Nasdaq-100QQQ$661.75−2.7%−2.6%−6.8%+8.1%+9.2%−13.8
Russell 2000IWM$272.95−1.7%−3.4%+0.9%+11.6%+16.1%−10.9
Global equity
MSCI ACWIACWI$153.00−2.0%−2.1%−3.0%+8.8%+10.9%−6.0
Developed ex-US (EAFE)EFA$102.64−1.5%−2.2%−0.9%+8.6%+11.4%−5.8
Emerging marketsEEM$58.89−3.3%−3.5%−6.3%+8.2%+9.2%−19.0
Rates
2-year Treasury par yield2y yield4.83%−5bp+27bp+9bp+136bp+121bp—
10-year Treasury par yield10y yield5.12%−3bp+24bp+3bp+94bp+98bp—
30-year Treasury par yield30y yield5.46%−1bp+20bp+3bp+62bp+72bp—
10-year TIPS real par yield10y real2.92%+1bp+26bp+18bp+99bp+106bp—
Credit
IG corporatesLQD$102.80±0.0%−1.8%−0.2%−3.0%−2.3%−4.8
High yieldHYG$76.73−0.3%−1.0%+0.4%±0.0%+1.5%−2.6
EM sovereign USDEMB$90.72−0.5%−2.0%−0.4%−1.5%−0.4%−4.3
Dollar
Nominal broad dollar index (H.10)USD broad (H.10)122.6513 Nov 2026+0.1%+0.6%+1.0%+2.4%+1.6%±0.0
US dollar bullish (UUP)UUP$29.56+0.2%+2.0%+2.2%+9.4%+8.8%±0.0
Euro (FXE)FXE$101.48−0.2%−2.0%−2.1%−5.7%−4.6%−8.3
Yen (FXY)FXY$57.46+0.1%−1.0%−0.9%−2.1%−3.5%−4.9
Commodities
GoldGLD$352.90−0.7%−7.0%−7.6%−11.0%−6.1%−34.0
SilverSLV$42.39−3.9%−11.9%−17.7%−34.2%−7.8%−91.3
WTI crude (USO)USO$140.03−3.8%−1.8%−13.4%+102.5%+96.2%−19.1
Brent crude (BNO)BNO$58.66−3.9%−0.5%−12.7%+107.1%+99.5%−19.6
CopperCPER$36.03−1.4%−6.2%−8.8%+3.1%+15.0%−14.0
Style spreads
Growth − valuelong − short, log points$127.3113 Nov 2026−1.1−1.2−13.7−23.5−25.4−36.0
Cyclicals − defensiveslong − short, log points$100.7713 Nov 2026−1.3−1.1−11.4−18.8−16.3−27.5
AI complex − defensiveslong − short, log points$44.8513 Nov 2026−3.2−0.4−28.9−25.8−27.6−48.7
About this exhibit
A table of the section's benchmarks: US and global equity funds, Treasury yields, credit, the dollar, commodity funds and style spreads. Each row gives the last price, the return or change over each window from a day to a year, and the distance from the 52-week high. Cells switch to z-score or percentile.
Markets at a glance
About two dozen benchmarks, data to 16 Nov 2026 unless a row names its own date: yields in basis points, style spreads in log points, the rest total returns
Source: Yahoo Finance; U.S. Treasury; Macro Review calculations (the pages linked)cross.overview-1

The year against cash

The S&P 500's total return each calendar year beside what cash earned

Exhibit 4 · Equities against cash, by calendar year

The S&P 500's total return each calendar year beside what the 3-month yield accrued; the last year is to date
Source: Yahoo Finance; U.S. Treasury; Macro Review calculations (the pages linked)cross.overview-6

The state of the section in six readings

One reading per stage, each from the page that owns it, with the date of the data behind it

Exhibit 5 · Headline readings from the Cross-Asset pages, with their data dates

Each stage's headline readings, in %, bp, a count or a state, each beside the date of its data; latest 16 Nov 2026
StageReadingValueData to
StateHow violently is the market moving?S&P 500 vol, 20-day23.2%16 Nov 2026
StateWhat is already priced?10-year above the hold path's fair value+106bp16 Nov 2026
DriversWhat regime is the market in?RegimeOnly the dollar hedges13 Nov 2026
DriversIs risk being paid for?Families reading risk-on313 Nov 2026
ExpectationsDo the accounts support it?After-tax profit share of GDP11.9%Q2 2026
Fair valueHow far is the index from the model's value?Gap to the model's value−41.5%16 Nov 2026
All 15 readings of the section
StageReadingValueData to
StateHow violently is the market moving?S&P 500 vol, 20-day23.2%16 Nov 2026
StateWho is crowded?Outside the 20th–80th1710 Nov 2026
StateWhat is protection costing?not published yet–
StateWhich way must dealers hedge?not published yet–
StateWhat is already priced?10-year above the hold path's fair value+106bp16 Nov 2026
DriversWhat regime is the market in?RegimeOnly the dollar hedges13 Nov 2026
DriversIs risk being paid for?Families reading risk-on313 Nov 2026
DriversDo the factors explain the index?Explained share, 63 sessions60%16 Nov 2026
DriversIs anything unnamed moving the sectors?Residual absorption ratio0.33816 Nov 2026
ExpectationsDo the accounts support it?After-tax profit share of GDP11.9%Q2 2026
ExpectationsHow fast is long-run growth running against the anchor?Long-run growth state, nominal GDP5.6%Q2 2026
ExpectationsHow much of a pump-price move reaches the CPI?Pump price +10% → headline CPI+0.45%Oct 2026
Fair valueWhat is the market paying?Cyclically adjusted P/E39.4Sep 2026
Fair valueHow far is the index from the model's value?Gap to the model's value−41.5%16 Nov 2026
OutlookWhich assets carry a model-conditional target?Assets with a model-conditional target213 Nov 2026
About this exhibit
A table with one or two readings per stage (State, Drivers, Expectations, Fair value), each with its value and data date, copied from the page that owns it; every reading of the section in a fold beneath.
The state of the section
Six readings, one or two per stage, each its page's own with the date of the data behind it
Source: Yahoo Finance; U.S. Treasury; Macro Review calculations (the pages linked)cross.overview-3

What changed

Each reading against the same reading on its own page's previous update

Exhibit 6 · What changed since each page's previous update

Each reading against the same reading on its own page's previous update
ReadingThenNowChange
How violently is the market moving?against 14 Nov 202622.323.2+0.9
Who is crowded?against 16 Nov 20261717±0
What is already priced?against 14 Nov 2026114106−8
What regime is the market in?against 13 Nov 2026Only the dollar hedgesOnly the dollar hedges–
Is risk being paid for?against 16 Nov 202633±0
Do the factors explain the index?against 16 Nov 20265960+1
Is anything unnamed moving the sectors?against 16 Nov 20260.3390.338−0.001
Do the accounts support it?against 16 Nov 202611.911.9±0.0
How fast is long-run growth running against the anchor?against 16 Nov 20265.65.6±0.0
How much of a pump-price move reaches the CPI?against 16 Nov 20260.450.45±0.00
What is the market paying?against 16 Nov 202640.739.4−1.3
How far is the index from the model's value?against 16 Nov 2026-42.8-41.5+1.3
Which assets carry a model-conditional target?against 16 Nov 202622±0
Linked views. 4 of the 10 thematic views with series read a page that carries a factor this section prices:
About this exhibit
What a row says is “the page said this last time and says that now”, which a reader can check against the page's earlier edition.
Source: Yahoo Finance; U.S. Treasury; Macro Review calculations (the pages linked)cross.overview-4

The years ranked, the whole board and the pages

Total return by calendar year, ranked; every reading of the section; every page, the data it runs to and its last update

Exhibit 7 · Benchmark total returns ranked within each calendar year, with cash

Total return, %, by calendar year since 2014, ranked within each year, with cash (the 3-month yield accrued) as a row
About
Rank2014201520162017201820192020202120222023202420252026 YTD
1VNQ+30BTC-USD+34BTC-USD+123BTC-USD+1,425Cash+2BTC-USD+92BTC-USD+303USO+65USO+29BTC-USD+154BTC-USD+122GLD+64USO+102
2TLT+27VNQ+2IWM+22EEM+37IEF+1USO+33GLD+25BTC-USD+60Cash+2SPY+26GLD+27EEM+34VNQ+12
3SPY+13IEF+2HYG+13EFA+25TIP−1SPY+31IWM+20VNQ+41GLD−1EFA+18SPY+25EFA+32IWM+12
4IEF+9SPY+1SPY+12SPY+22TLT−2VNQ+29SPY+18SPY+29HYG−11IWM+17USO+13SPY+18SPY+9
5LQD+8Cash±0EEM+11IWM+15GLD−2IWM+25TLT+18IWM+15TIP−12GLD+13IWM+11IWM+13EFA+9
6IWM+5EFA−1VNQ+9GLD+13HYG−2EFA+22EEM+17EFA+11EFA−14VNQ+12HYG+8HYG+9EEM+8
7TIP+4LQD−1GLD+8TLT+9LQD−4EEM+18LQD+11TIP+6IEF−15HYG+12EEM+6IEF+8Cash+3
8HYG+2TIP−2USO+7LQD+7SPY−5GLD+18TIP+11HYG+4LQD−18LQD+9Cash+5LQD+8HYG±0
9Cash±0TLT−2LQD+6HYG+6VNQ−6LQD+17IEF+10Cash±0SPY−18EEM+9VNQ+5TIP+7TIP−1
10GLD−2IWM−4TIP+5VNQ+5IWM−11TLT+14EFA+8LQD−2IWM−20Cash+5EFA+3TLT+4IEF−3
11EEM−4HYG−5EFA+1TIP+3EFA−14HYG+14HYG+4IEF−3EEM−21TIP+4TIP+2Cash+4LQD−3
12EFA−6GLD−11TLT+1IEF+3EEM−15TIP+8Cash±0EEM−4VNQ−26IEF+4LQD+1VNQ+3TLT−5
13USO−42EEM−16IEF+1USO+2USO−20IEF+8VNQ−5GLD−4TLT−31TLT+3IEF−1BTC-USD−6GLD−11
14USO−46Cash±0Cash+1BTC-USD−74Cash+2USO−68TLT−5BTC-USD−64USO−5TLT−8USO−8BTC-USD−22

Total returns from adjusted closes, measured from the last close of the previous year; a year a fund did not yet trade in is absent rather than partial. Hover a square for the asset and its return.

About this exhibit
A quilt of calendar-year total returns for the section's benchmark funds and bitcoin, ranked within each year, with cash (the 3-month yield accrued) as its own row; the last year is to date.
The year, and the years before it
Total return by calendar year since 2014, ranked within each year, with cash (the 3-month yield accrued) as a row
A year in which everything lost money is a different year from one in which everything lost money while cash earned five per cent.
Source: Yahoo Finance; U.S. Treasury; Macro Review calculations (the pages linked)cross.overview-2

Exhibit 8 · Pages and their status

The 25 pages of the section in map order: the question each answers, the data it runs to and its last update
PageIts questionStatusData toUpdatedChecks
State
PerformanceWhat has moved across assets, and how unusual is each move against its own volatility and history? live16 Nov 202617 Nov 202623
VolatilityHow volatile is each market in its own history, is volatility itself unstable, and what do FOMC weeks do to volatility products? live16 Nov 202617 Nov 202628
PositioningWho is crowded in rates, equities, currencies and commodities — and how far is that from their own history? live10 Nov 202617 Nov 202610
What is pricedHow much of the return has the rates path already taken, and how much is still to come? live16 Nov 202617 Nov 2026109
Volatility & positioning
Implied volatility– planned–––
Skew & term structureWhich tail of each market is bid, and how does the price of that insurance change with the horizon? live16 Nov 202617 Nov 202666
Dealer positioningWhere are dealers long and short gamma in each market, and is that position moving? live16 Nov 202617 Nov 202655
Drivers
Risk monitorWhere does risk appetite stand, in every family the industry uses? live13 Nov 202617 Nov 202649
Correlations & regimeWhich co-movement regime are markets in, how sure is the vote, how have assets moved since it began, and what does the macro say? live13 Nov 202614 Nov 202634
Risk sentimentIs the market paying for risk, and what does risk cost to hold? live13 Nov 202617 Nov 202657
Return analysis
Sources of returnWhat is each asset's return made of: income, growth or a change in what the market pays for it? live16 Nov 202617 Nov 202641
Factor analysisWhich factors drove this asset's return, and how exposed is it to each? live16 Nov 202617 Nov 202671
Fundamentals
EarningsWhat does the index earn, and what supports it? liveApr 202617 Nov 202692
Payout & reinvestmentWhere do the index's earnings go? liveApr 202617 Nov 202638
Profitability & balance sheetWhat return does the index earn on its capital, and on how much leverage? liveApr 202617 Nov 202639
CompositionWhose earnings are the index's earnings? liveApr 202613 Nov 202620
Long-run growthWhat terminal growth does the valuation assume, and is it consistent with the data? liveApr 202617 Nov 202676
Fair value
ValuationWhat is the S&P 500 worth, and what premium is the market paying for equity risk? liveSep 202617 Nov 2026278
Statements valueWhat is the index worth on its own statements, and what premium does today's price imply? live16 Nov 202617 Nov 202667
Fair value boardWhat is each macro asset worth, and how much should each model be trusted? live16 Nov 202617 Nov 2026109
Valuation historyHas this valuation gap predicted ten-year returns? liveNov 202617 Nov 202636
Sectors & styles
Sector earningsHow far are each basket's forward earnings above their own five-year average, and what would carry them? live13 Nov 202617 Nov 202633
Sector valuationHow far is each sector's forward P/E from its own normal, and does that gap carry information about next year's multiple? liveNov 202617 Nov 202667
Growth against valueIs growth dear against value, and has the market paid for beta while the ratio moved? live13 Nov 202617 Nov 202614
Method and recordHow is each number made, and how far can each model be trusted? live13 Nov 202617 Nov 20267
Source: Yahoo Finance; U.S. Treasury; Macro Review calculations (the pages linked)cross.overview-7

Appendix

Where each number comes from, what is computed on this page and how the page is held to its sources171 checks, all pass
About this page

One screen to start the Cross-Asset section from. Every number on this page is another page's and is compared with it on every update; the one exception is the returns quilt, whose year to date is checked against the Performance page's.

Exhibit 9 · Definitions

  • Where the numbers come from

    Almost nothing here is computed. The rows of the glance are the Performance page's own, every reading of the board is its page's own, and every reading the takeaways and the tile row are written from is a value another page publishes — each taken with its data date and compared with that page on every update, which is what the checks list. The S&P 500's model value and its words (“the model's value … · record unproven · no target”) are the Fair value pages' one register, printed as they print it.

  • The quilt

    The one computation: total returns by calendar year since 2014 from adjusted closes, ranked within each year, with cash accrued from the 3-month par yield as a row of its own. A check asserts this page's year to date equals the performance board's for every fund both carry. Calendar-year returns are measured from the last close of the previous year, so a year a fund did not yet trade in is absent rather than partial.

  • The gap's history

    The hero is the Valuation history page's own monthly series of the gap to fair value on the data available at each date (real-time), over the last twenty years; the title's run is the months back to the last one on the other side of zero. The Valuation page's daily gap and the Fair value board's are later readings of the same model, each with its own date.

  • The board

    One reading per page, each its page's own tile, with the date of the data behind it. The section's pages end on different days — prices on the last session, the attribution and the gauge on the last full week, the curve behind What is priced on the Treasury's latest release, positioning on the CFTC's Tuesday — so a reading carries its own date rather than the page's. A reading whose page is not published yet says so and carries no number. The six-row table under the takeaways is six of these readings, one or two per stage; the whole board is in the Detail fold, in the order of the section's blocks.

  • Dates

    Where two pages of the section print the same figure to different dates — the S&P 500's year to date on the performance board and on the attribution page — the takeaways print both, each with its end date, so the two numbers read as two dates rather than as a disagreement.

  • The flow strip

    The flow strip at the foot of this page, the US Macro and Thematic overviews and Home prints one reading under each stage — State, Drivers, Expectations, Transmission, Scenarios, Outlook — each its owning page's own tile with its date; the other three pages take these six from this one and compare them with it, so the four strips cannot disagree.

  • What changed

    Each reading against the same tile on the previous update of its own page — not against a recomputation. What the row says is “the page said X last time and says Y now”, which a reader can check against the page's earlier edition. A tile whose label carries its reading (the regime's state name) is matched by the label's stem, and the row names the earlier label.

  • Linked views

    A view is linked to this section by a factor, not by a page: the factor map names the factors the Cross-Asset pages price (oil, rates, credit, breakeven inflation) and the pages that carry each. A view is linked when a page one of its series reads is in the map; a series a view page computes itself from a publisher is named as the view and the series. 4 of the 10 views with series are linked on this update.

  • Pages and their status

    The directory of the section, in a Detail fold: its 25 pages in map order, each with the question it answers, the data it runs to and the day it was last updated — the rows Home lists, for one section. The board has 13 of its 15 readings published.

  • Licence

    Prices are Yahoo Finance data, shown with their attribution; Treasury yields are the Treasury's. Every other number is another page's, with that page's own sources behind it.

Source: Yahoo Finance; U.S. Treasury; Macro Review calculations (the pages linked)cross.overview-5

Before you read

Macro Review is a personal research site, still being built.

Nothing on it is investment advice, a recommendation or a solicitation.

Check every figure for yourself. Each names its publisher — confirm it there before you rely on it.

Figures can be revised, delayed or wrong.

Cones, fans, scenarios, fair values and “what followed” tables are conditional on the model each page names: they describe, they do not predict.

No warranty is given; use is at your own risk.