Snowshagal · INVESTMENT NOTE
NO.05 · 2026.09.24
Snowshagal ReportUS INFLATION · BEA 2026 ANNUAL UPDATE
INVESTMENT NOTENO.052026.09.24
09.30 · August PCE and annual update released together

Inflation Getsa New Ruler

Calculation methods change for three core PCE components · the main revision window is recalculated and released the same day

Watch not for a break in the time series, but for the size of the revision · and how far that revision differs from the published benchmark estimate of 10~20bp

July 2026 core PCE Old ruler 0 1 2 3 3.34 New ruler 0 1 2 3 3.15 Same reality · different scale · different number
Concept illustration · numbers on the rulers are Hanwha Investment & Securities' reconstructed July core PCE YoY readings · not official BEA figures
Release · Korea time
09.3021:30
Same time as August PCE · revised monthly time series released simultaneously1
Share of affected components
4.16%
Share of total PCE · Hanwha Investment & Securities calculation · not official BEA weights2
Official BEA effect
Not disclosed
Preview gives no numerical estimate of the size or sign of the effect3
Published pre-release estimates from major institutions
10~20bp
Estimated downward revision to core PCE · GS 20 · JPM 10 · June estimates based on May data · not market pricing4

The official time series will not break · the risk is mixing pre-revision and post-revision figures · the new information on the methodology change is the gap versus benchmark estimates

Information cutoff · 2026.09.24 20:00 KST · excludes releases and overseas intraday data after that time · all old-vs-new method comparisons are private-sector estimates

snowshagal.com

AT A GLANCE

WHEN
Sep. 30 (Wed) 21:30 KST · same time as August PCE
WHAT
New calculation methods for three core PCE components · main revision window 2021 Q1-2026 Q1
HOW MUCH
Official BEA effect not disclosed · Hanwha reconstruction of July core PCE −0.19%p
BENCHMARK ESTIMATES
GS about 20bp · JPM about 10bp lower · published institutional estimates · based on May data
RISK
Splicing the pre-revision archive to post-revision figures without labeling the break in vintage
WATCH
Gap versus benchmark estimates · monthly path · simultaneous releases including the third estimate of GDP
PAGE 01RELEASE STRUCTURE
Two numbers · one timestamp

Same day,
two numbers

New August reading + recalculation of the main revision window · released at the same timestamp

At the end of its July Personal Income and Outlays release, the BEA set the next release for September 30 at 8:30 a.m. EDT: Personal Income and Outlays, August 2026. The same release also states that the annual updates to national, industry, and regional statistics will begin on the same day for the first time1 · 21:30 KST on September 30

With the next release, today’s data will be supersededThe current figures will be superseded with the next release
Updated monthly estimates of personal income and outlays will be released on September 30, along with the estimate for August 2026.Updated monthly estimates will be released together with the August estimate1

Not a break - a full rewrite

Same month, two vintagesConcept
Pre-revision archive · through July September 30 revised vintage · through August 20212023202526.8 The fake drop created by mixing vintages Compare like vintage with like vintage
Line shapes and spacing are illustrative · not actual PCE values · splicing July (pre-revision) directly to August (post-revision) can make a change in the ruler look like a fall in inflation
09.30WED · 21:30 KST
BEA August PCE + annual updateAugust new reading and revised monthly time series · third estimate of Q2 GDP released the same day1, 3, 6
10.22THU
Bank of Korea Monetary Policy BoardBetween the September 30 release and the October FOMC7
10.27~28FOMC · no SEP
First meeting after the revisionFirst meeting held with the new inflation data in hand8
12.08~09FOMC · SEP released
First projections after the revisionFirst Summary of Economic Projections produced with the new ruler8
PAGE 02AFFECTED COMPONENTS
Three line items

Three components,
4.16% of PCE

All are outside food and energy · all feed directly into core PCE

The BEA's June Survey of Current BusinessSurvey of Current Business annual-update preview explicitly lists all three changes3 · the shares below are not official BEA weights, but PCE shares calculated by Hanwha Investment & Securities · this note uses total PCE as the denominator throughout2

Pieces of total PCE affected by the changesShare of total PCE · %
Total PCE 100% 4.16% 2.11 1.23 0.82 Portfolio management Software Legal
Hanwha Investment & Securities calculation · not official BEA weights · lower bar magnifies the end segment of the upper bar

Portfolio management & investment advice2.11%

OLDNominal spending / PPI sensitive to asset prices
NEWExtrapolate real quantity using BLS employment data (CES) as the basis
Entire main revision window · from 2021 Q1

Computer software & accessories1.23%

OLDCPI-based
NEWComposite index: CPI + PPI for game software publishing + hosting PPI
Entire main revision window · from 2021 Q1

Legal services0.82%

OLDCPI
NEWBEA composite index based on detailed PPIs for legal services consumed by households
Beginning in 2024 · source wording: "beginning in 2024"
PAGE 03MECHANISM
Where stocks leaked into prices

The channel through which
stocks leaked into inflation

The change does not swap the divisor; it measures quantity separately

Conceptually, the biggest change is portfolio management · asset-management fees are typically proportional to assets under management · when stocks rise, fees on the same service automatically rise, and under the old method that movement is absorbed into prices2

OLD METHODBased on Hanwha's explanation
Total nominal fees+18.5%June 2026, YoY
/ divide by an asset-price-linked PPI
Real quantityabout -3.1%Residual implied after deflating by the PPI
Stock-market movement remains in the price measure
Price inflationabout 22%
NEW METHODBEA preview · CES-based quantity extrapolation
Total nominal fees+18.5%Same starting point
Measure quantity first
Real quantity · extrapolated from industry employment+4.7%CES industry employment growth
Only the residual goes to price
Price inflationabout 13.2%

June 2026, YoY · Hanwha Investment & Securities reconstruction · quantity under the old method is not a BEA-published figure but the residual implied by PPI deflation2, 3 · Deflatoris not simply swapped out; instead the methodology changes direction through quantity extrapolation

S&P 500 move of 1% → price responsepercentage points · Hanwha Investment & Securities
00.250.500.751.0 New 0.43 Old 0.67 Reduced portion
Sensitivity
−36%

Sensitivity 0.67%p → 0.43%p2 · the response remains at roughly two-thirds of its former level · it does not disappear

FED · STEPHEN MIRAN · 2025.12.15Then-Fed Governor · official speech
What ought to be recorded as an increased quantity of services consumed is instead recorded as increased prices.

The argument: part of the higher management fees generated by rising equity prices should be recorded as greater service quantity, not higher service prices9

The CES-based quantity extrapolation introduced in the BEA annual update moves directly toward resolving this price-versus-quantity classification problem3

The equity-market-to-PCE-price link is weakened, not severedA structural channel through which asset-price moves had seeped into portfolio-management PCE prices · the revision narrows that channel
Hanwha simulation · 2009 lies outside the official revision window
EpisodeOld methodNew method
Mar. 2009 financial crisis−0.26+0.05
Apr. 2021 equity surge+0.48+0.22

Contribution of portfolio management to core PCE · percentage points · Hanwha calculations show narrower upside and downside swings · because the main BEA revision window is 2021 Q1-2026 Q1, the 2009 figure is a historical simulation applying the new method backward · the April 2021 figure is also not an official BEA result2, 3

PAGE 03 · SUMMARY
  1. Old: deflate nominal spending with an asset-price-linked PPI → equity moves are absorbed into prices
  2. New: measure quantity first using employment data → price is the residual
  3. Result: equity-to-PCE-price sensitivity falls by roughly one-third · the link itself remains
PAGE 04WEIGHT OF THE NUMBERS
The heavy digit

The largest downward
revision is in software

Share 1.23% · adjustment -0.14 percentage point

The narrative weight is in portfolio management; the numerical weight is in software · component-level old-vs-new differences for July 20262

July 2026 core PCE YoY · adjustment by componentpercentage points · Hanwha reconstruction
SoftwarePCE share 1.23%
−0.14
Portfolio managementPCE share 2.11%
−0.07
Legal servicesPCE share 0.82%
+0.02
−0.15−0.10−0.050+0.05
Total of three components · core PCE YoY 3.34% → 3.15%
Hanwha reconstruction · official BEA July reading was 3.3%−0.19%p
Zero is the vertical line · left is downward, right is upward · all figures are Hanwha Investment & Securities calculations, not official BEA recomputations · July will also be replaced in the revised monthly series on September 30, with no guarantee the official revision will match these estimates1, 2

Why a component with 60% of the weight delivers twice the adjustment

The existing software index itself shows an extreme increase: 70.3 in July 2025 → 85.2 in July 20262

Current indexCPI-based
+21.2%
New composite indexCPI + two PPI series
+8.6%
FEDS NOTE · 2026.05.22Barbarino · Diercks · Miran

Fed researchers had already flagged the same measurement problem · PCE software prices from November 2025 through March 2026 rose at 73% annualized · more than 9 standard deviations10

  • Category mismatch · PCE software is centered on software publishing, while the CPI used to deflate it includes flash drives and blank media
  • They also flag gaps in measuring prices and quality adjustment for SaaS and AI-bundled products

FEDS Notes reflect the authors' views, not the official position of the Federal Reserve · an independent source identifying the same anomaly as the Hanwha reconstruction

PAGE 03 uses June 2026 figures; this page uses July · the old-vs-new gap in portfolio management varies materially by month, so figures from the two pages cannot simply be multiplied together

×30+ Core CPI 0.035% Core PCE 1.2% range

Same component, more than 30 times the weight

Software weight · core CPI 0.035% vs core PCE in the 1.2% range2 · a leading point where CPI and PCE diverge despite measuring the same broad concept of inflation · JPMorgan's Abiel Reinhart likewise notes that PCE gives the category more than 30 times the weight, while the CPI-priced items are not conceptually identical to the PCE definition4

PAGE 05POLICY INTERPRETATION
From measurement to policy

How Much Does It
Matter for Policy?

YoY revision size ≠ monthly policy signal
September SEP · median core PCEQ4/Q4 · %
2% target 3.42.52.2 2.0 2026202720282029 SeptemberJune
  • September 16 FOMC Summary of Economic Projections · prepared two weeks before the revision release 11
  • Core PCE: 3.4% in 2026 (3.3% in June) → 2.0% in 2029 · the return to 2% is pushed to the end of the projection horizon
  • Median federal funds rate for end-2026: 3.8% → 4.1%
  • How much participants incorporated the coming revision is not disclosed · the BEA had previewed the changes twice, in June and August, so it is difficult to argue they were unknown3, 5

①Do not divide the YoY gap by 12

In its own December-hike scenario, Yuanta Securities uses monthly core PCE inflation of 0.25-0.30% as consistent with a hike; two consecutive months at 0.2% or less as leaving room to hold. This is a securities-firm scenario, not a Federal Reserve rule12

The temptation is to divide Hanwha's estimated -0.19 percentage-point YoY revision by 12 · that calculation is invalid · the YoY gap is the accumulated difference across 12 monthly changes, not a representative monthly value13

What the aggregate figure tells you · and what it does notHanwha reconstruction · YoY
2025
Annual result
0.00%p · no difference
Monthly path
Cannot be determined · no published monthly values
2026 · Jan.-Jul.
Jan.-Jul. result
−0.13%p
Average monthly gap
−0.12%p
Monthly distribution
To be checked in the revised series on September 30
Aggregate figureBack-solving the monthly path
The source discloses only annual aggregates · "no difference" for 2025 means the annual result was unchanged, not that every month was identical · the gap in an individual month could be much larger than the average or even have the opposite sign · the monthly path can only be checked once the revised monthly series is released on September 302, 13

②Known · estimated · unknown

This is not a secret change · major institutions began estimating the downward revision immediately after the BEA's June preview · according to Reuters on June 29, 2026, Goldman Sachs estimated May core PCE would be revised from 3.4% to 3.2% (about 20bp), while JPMorgan estimated 3.3% (about 10bp) 4 · Axios likewise summarized analysts' broad expectation as roughly a 0.2 percentage-point downward revision14

Estimated downward revision to core PCE YoY · published institutional mapbp
10JPMorgan
20Goldman Sachs
19Hanwha · July
0102030
Below 10Smaller decline than benchmark estimates
10~20Published benchmark range
Above 20Larger decline than benchmark estimates
Estimates use different reference months · GS and JPM are based on May core PCE, Hanwha on July · none are official BEA figures · this is an overlay of published institutional estimates, not a consensus survey or market pricing2, 4
KNOWN

Already known

  • Methodology changes for three components
  • Main revision window · 2021 Q1-2026 Q1
  • Release on September 30
ESTIMATE

Published estimates from major institutions

  • Major institutional estimates point downward
  • GS about 20bp · JPM about 10bp
  • Hanwha July reconstruction 19bp
UNKNOWN

Only known on September 30

  • Official BEA revision size and sign
  • Monthly path
  • August underlying inflation
  • Effect of newly incorporated source data
First comparison on September 30 · benchmark vs actual revision gapAssess the methodology contribution only after reviewing the BEA revision materials · "core PCE revised down 0.2 percentage point" alone is not new information · the total revision also includes newly incorporated source data, so even an actual 25bp revision against a 20bp estimate cannot immediately be read as a "+5bp methodology surprise"3, 5, 6
Easy to miss · the basis of August forecasts
  • How much an August forecast incorporates the new methodology may differ by institution13
  • There is also no basis for assuming major institutions simply kept using the old method after the BEA disclosed the changes in June
  • Do not interpret the gap between consensus and actual as entirely an underlying-inflation surprise
③Not an input to mechanically plug into a neutral-rate formula
  • Kiwoom Securities' nominal neutral rate of 4.20% assumes core PCE at 2.5% · using 2.0% yields 3.735%15
  • Difference between the two points: 4.20-2.5=1.70 · 3.735-2.0=1.735 · residual 0.035 percentage point → effectively nominal neutral rate = real neutral rate + inflation assumption 13
  • The 2.5% figure is a long-run inflation assumption, not a current-month observation → there is no reason to mechanically lower it solely because of this methodology change
  • Hanwha's reconstructed cumulative inflation differs by only 0.1 percentage point · Jan. 2021-Jul. 2026: current 22.9% vs revised 23.0%2
  • The U.S. 10-year yield, however, reflects the policy path, inflation surprises, andterm premiumsimultaneously · an unexpectedly large revision could therefore move market yields
PAGE 05 · SUMMARY
  1. YoY revision size ≠ monthly signal · do not divide by 12
  2. The first September 30 comparison is published benchmark estimates of 10-20bp versus the actual total revision · determine the methodology contribution only after checking BEA revision materials
  3. The degree to which August forecasts embed the new methodology may vary by institution · do not read the entire forecast miss as an underlying-inflation surprise
PAGE 06COUNTERPOINTS & CHECKPOINTS
Counterpoints · checkpoints

Evidence on the other side and
what to verify

Hanwha reconstruction · cumulative difference is small and the sign changes across episodes
01

Episode effects matter more than the cumulative level

Jan. 2021-Jul. 2026 cumulative core PCE inflation: current 22.9% · revised 23.0% · only a 0.1 percentage-point gap over five years and seven months · in Hanwha's reconstruction, episode-level differences are larger than the long-run cumulative-level difference · not an official BEA result2

Current22.9%
Revised23.0%
02

The sign changes by year and period

−0.10 +0.05~0.13 0 −0.13 202122~2420252026

Average monthly gap: -0.12 percentage point in Jan.-Jul. 2026, +0.01 percentage point over the full sample · all on a YoY basis · not a one-way downward revision2

03

The numbers themselves are still estimates

All adjustment figures in the main text are Hanwha Investment & Securities reconstructions · because the BEA Fisher chain-type index cannot be replicated from the available inputs, the author approximates it with monthly expenditure-share-weighted aggregation · the author also states that legal-services and software sub-indexes are arbitrarily equal-weighted at one-third each2

04

An annual update changes more than methodology

The same update also incorporates newly available source data · the difference between old and new values cannot be attributed entirely to the three calculation-method changes3, 5

05

Energy's direct effect is mainly in headline PCE

July headline PCE was +3.7% YoY vs +3.3% core · the methodology changes apply to core components, while the direct composition effect of energy enters headline inflation · energy costs can still pass indirectly into services and transportation · the Fed's formal inflation objective is headline PCE, while core is a signal of the underlying trend1

06

A credibility issue separate from technical validity

Axios summarized analysts' broad expectation as about a 0.2 percentage-point downward revision, judging the changes technically defensible while also noting that the optics are poor at a time when statistical-agency independence is under attack14

Four things to check on the night of September 30

WATCH 01

Compare monthly series before and after revision

Compare the pre-revision archive with the new monthly series to establish the total revision · then check the NIPA revision table and SCB annual-update article released the same day for explanations of new source data and methodology changes1, 6

WATCH 02

Distance from the 10-20bp benchmark estimates

The first comparison is the gap between published institutional estimates and the actual total revision4 · use the revision table and SCB annual-update article together to determine how far the effects of new source data and methodology can be decomposed · because new August PCE, the third estimate of Q2 GDP, and other releases arrive at the same time, do not attribute the entire intraday yield move to the methodology revision6

WATCH 03

Concentration in the monthly path

Which months account for the YoY gap · the revised monthly series is needed to identify month-specific gaps · only then can monthly inflation-rate thresholds be evaluated2, 13

WATCH 04

Compare the December SEP with September's 3.4%

The December 9 SEP is the first set of projections after the revision · the difference between the two includes both methodology effects and actual inflation developments · do not read the full difference as a forecast downgrade8, 11

Transmission to Korea
  • Not a direct channel · if changes in U.S. inflation alter expectations for the Fed policy path, they can pass through U.S. long-term yields into Korean long-term yields
  • A recent Bank of Korea study (BOK Economic Research 2026-15) also finds that in Korea-U.S. long-term yield comovement, global inflation shocksmake the largest contribution (41.0%), while U.S.-origin shocks are transmitted mainly through the expected future monetary-policy pathrather than risk compensation16
  • Accordingly, the methodology change alone cannot be used to mechanically calculate a decline in Korean government-bond yields (PAGE 05, item 3)
  • The most common mistake will be to treat a lower core PCE reading as an underlying-inflation slowdown and mechanically reduce the probability of a December hike · the opposite mistake is to dismiss the entire move as methodology · the two can only be separated after comparing the pre- and post-revision series
  • For Korea, the domestic real-economy variables are the end of share buybacks and trading value discussed in NOTE NO.047
THE FED'S CURRENT FRAME
Chair Warsh · Jackson Hole · 2026.08.28"Nor should we rely on isolated data points. Trends matter most." · he also argues that disaggregating the 199 individual components of the PCE price measure is instructive17

Why one lower number should not be read as an immediate easing signal · the speech does not directly discuss the BEA methodology change

CLOSING NOTE
An indicator is not a fact but
a measurement convention.

Part of a lower number may reflect a change in the ruler, not a change in inflation itself.

This note is a pre-release check for a day when an indicator's definition changes · it is not a forecast of asset direction or a trading recommendation · all old-vs-new comparisons are private-sector reconstructions, not official BEA recomputations · the BEA preview provides no numerical estimate of the size or sign of the change · the 10-20bp benchmark estimates are based on reporting dated June 29, 2026, and are neither consensus polling nor market pricing · Fed projections are from the September 16, 2026 SEP; annual values are Q4/Q4 · observed inflation is July 2026 (released August 26), YoY · the actual September 30 release may differ from the estimates in this note

GLOSSARYGLOSSARY
PCE price index
Personal Consumption Expenditures price index · the Fed's formal inflation objective is headline PCE; core PCE is used to read the underlying trend
Core PCE
PCE excluding food and energy · closer to the portion of inflation monetary policy can influence
BEA
U.S. Bureau of Economic Analysis · the agency that produces GDP and PCE statistics
Annual update
Once a year, historical estimates are recalculated using new source data and updated methods · methodology changes and newly incorporated source data enter together
Deflator
A price index used to convert nominal spending into real spending · the choice of index determines how the movement is split between prices and real quantity
Quantity extrapolation
A method that estimates real quantity directly from another data source such as employment instead of deriving it by dividing nominal spending by a price index
vintage
A version of the same statistic as published at a particular point in time · pre-revision archives and post-revision values are different vintages
SEP
Summary of Economic Projections by FOMC participants · includes the dot plot · released at the March, June, September, and December meetings · annual values are Q4/Q4
Neutral rate
The interest rate consistent with neither stimulating nor restraining the economy · an estimate that depends in part on the assumed long-run inflation rate
Term premium
Additional compensation required to hold long-duration bonds · helps explain long-term yields alongside the neutral rate
SOURCES (17)

Information cutoff · 2026.09.24 20:00 KST · excludes releases and overseas intraday data after that time

1. U.S. Bureau of Economic Analysis · Personal Income and Outlays, July 2026(released 2026.08.26) · July PCE +0.2% MoM, +3.7% YoY · core +0.2% MoM, +3.3% YoY · “Next release: September 30, 2026, at 8:30 a.m. EDT, Personal Income and Outlays, August 2026” · “With the next release, today’s data will be superseded” · “Updated monthly estimates of personal income and outlays will be released on September 30, along with the estimate for August 2026.” · “the 2026 annual updates of national, industry, and regional data will begin on the same day for the first time: September 30, 2026” · explicitly states that monthly personal income and outlays are included in the annual update · KST conversion uses EDT (UTC-4): 09.30 21:30 KST.
2. Hanwha Investment & Securities, "STRATEGY NOTE · Assessing the Impact of Changes to PCE Calculation Methods" report dated 2026.09.21 · posted on website 09.22(Kim Sung-soo) · affected-component shares of total PCE: portfolio management & investment advice 2.11% · legal services 0.82% · computer software & accessories 1.23% (calculated by the author, not official BEA weights) · July 2026 core PCE 3.34% → 3.15% · component adjustments: software -0.14 percentage point · portfolio management -0.07 · legal services +0.02 · portfolio management, June 2026 YoY: total fees +18.5%, industry employment +4.7%, new-method price about 13.2% vs old about 22% (real quantity about -3.1% is the residual implied by PPI deflation) · price response to a 1% move in the S&P 500: 0.67 percentage point → 0.43 · March 2009 contribution -0.26 → +0.05 · April 2021 +0.48 → +0.22 (2009 is a historical simulation outside the BEA's official revision window) · current software index 70.3 in 2025.07 → 85.2 in 2026.07 (+21.2%), new method +8.6% · core CPI weight 0.035% vs core PCE weight 1.2% · cumulative Jan. 2021-Jul. 2026: current 22.9% vs revised 23.0%; average monthly gap over the full period +0.01 percentage point · annual gaps: 2021 -0.10, 2022-2024 +0.05 to +0.13, 2025 no difference, 2026 Jan.-Jul. -0.13; 2026 monthly average -0.12 · an in-house reconstruction: the author states that the BEA Fisher chain-type index could not be replicated, so monthly expenditure-share-weighted aggregation was used as an approximation; legal-services and software sub-indexes were arbitrarily equal-weighted at one-third each · because the same report cites software at 1.23% of total PCE and about 1.2% of core PCE, with different denominators, this note uses total PCE shares consistently in its tables and text and describes the CPI comparison only as "more than 30 times."
3. U.S. Bureau of Economic Analysis · Preview of the 2026 Annual Update of the National Economic Accounts, Survey of Current Business June 2026 · portfolio management: "The quantity of portfolio management services will be derived using a U.S. Bureau of Labor Statistics (BLS) Current Employment Statistics (CES)-based quantity extrapolator" · legal services: "A BEA composite price index derived using detailed PPIs for selected legal services consumed by households will replace the consumer price index (CPI) for legal services beginning in 2024" · software: "A BEA-composite price index derived using the CPI for computer software and accessories; the PPI for game software publishing; and the PPI for hosting, ASP and other IT infrastructure provisioning services" · retrospective recomputation: "first quarter of 2021 through the first quarter of 2026" · the annual update includes methodology changes together with “more complete source data” incorporation of more complete source data · the BEA preview provides no numerical estimate of the size of the effects.
4. Reuters · Methodology changes could lower US May core PCE inflation, economists say 2026.06.29(Lucia Mutikani) · Goldman Sachs estimated May 2026 core PCE YoY would be revised from 3.4% to 3.2%, while JPMorgan estimated 3.3% · JPMorgan's Abiel Reinhart: "the PCE index receives more than 30 times as much weight as the CPI index, while the items priced for the CPI index are not conceptually identical to the PCE definition" · the same report was also carried by Investing.com(Jaiveer Shekhawat) also published it the same day · estimates refer to May core PCE · these are estimates from two institutions only; there is no evidence that 10-20bp was embedded in a consensus survey or market pricing
5. U.S. Bureau of Economic Analysis · Annual Update of GDP, Industry, and State Stats Publicly Available Starting Sept. 30 2026.08.17 · reconfirms that the annual update includes improvements to "inflation-adjusted consumer spending statistics and the personal consumption expenditures (PCE) price index" · cites both "more complete source data" and "methodological improvements" as purposes of the update · revision window for most statistics is 2021 Q1-2026 Q1.
6. U.S. Bureau of Economic Analysis · Information on 2026 Annual Updates to the National, Industry, State, and County Statistics · the September 30 release list includes “News releases for 2nd Quarter 2026 GDP (Third Estimate) and August Personal Income and Outlays” · all monthly and quarterly NIPA tables and industry economic-accounts tables · “Table of NIPA Revisions: Components Detail and Major Source Data and Conceptual Changes Incorporated, 2021–2025” · SCB articles “The 2026 Annual Update of the National Economic Accounts”and “Updated Summary of NIPA Methodologies” are listed for simultaneous release · the page does not separately state a release time, so this note uses the 08:30 EDT time in Source 1
7. Bank of Korea · Monetary Policy Decision Meeting Schedule and Materials · eight monetary-policy decision meetings in 2026 (Jan., Feb., Apr., May, Jul., Aug., Oct., Nov.) · next meeting: Thursday, October 22 / for the domestic liquidity section, see Snowshagal INVESTMENT NOTE NO.04 (2026.09.19).
8. Federal Reserve Board · FOMC Calendars, Statements, and Minutes · eight scheduled meetings in 2026 · October 27-28 (no SEP) · December 8-9 (SEP released) · SEP is released at four 2026 meetings: March, June, September, and December.
9. Federal Reserve Board · Governor Stephen I. Miran, The Inflation Outlook 2025.12.15 (Columbia University SIPA) · on portfolio-management PCE: “What ought to be recorded as an increased quantity of services consumed is instead recorded as increased prices.” · "The PCE portfolio management measure contains no signal regarding tightness in the asset management industry, let alone the economy." · "It contains no signal for the inflation forecast going forward." · remarks by a then-serving Federal Reserve governor, before the BEA announced the 2026 change
10. Federal Reserve Board · FEDS Notes, Alessandro Barbarino · Anthony M. Diercks · Stephen Miran, Measurement of “Computer Software and Accessories” Inflation 2026.05.22 · prices in the category from November 2025 through March 2026 posted “a record 73 percent annualized increase” · relative to the historical average, “more than 9 standard deviations” · category mismatch: flash drives and blank media carry zero PCE weight but are included in the CPI category · also discusses measurement and quality-adjustment issues for SaaS and AI-bundled products · FEDS Notes express the authors' views and are not an official position of the Federal Reserve
11. Federal Reserve Board · Summary of Economic Projections, September 16, 2026 · median core PCE: 3.4% in 2026 (3.3% in June), 2.5% in 2027, 2.2% in 2028 (2.1% in June), 2.0% in 2029 · median PCE: 3.7% in 2026 (3.6% in June) · median federal funds rate: 4.1% in 2026 (3.8% in June) · unemployment 4.1% in 2026-2029 · real GDP 2.3% in 2026 · SEP annual values are Q4/Q4 and are not directly comparable with monthly YoY inflation readings.
12. Yuanta Securities, "September FOMC Review · Kevin Warsh Did Not Define the Range," 2026.09.17 · forecasts an additional 25bp hike in December and a hold in October · in its own scenario, uses monthly core PCE inflation of 0.25-0.30% as consistent with a hike, while two consecutive months at 0.2% or less would leave room to hold · this is the securities firm's December-hike scenario, not a threshold announced by the Federal Reserve · forecasts 2-year Treasury at 4.65-4.75% and the 10-year near 5.00%.
13. 13. Calculations and judgments in this note · the calculation that divides the YoY gap by 12 to convert it into a monthly signal was not used · the YoY gap is the cumulative result of 12 monthly differences and therefore is not a representative value for each month; Source 2 itself shows time concentration, with no 2025 annual difference but a -0.12 percentage-point average monthly gap in Jan.-Jul. 2026 · the identity check takes the two points in Source 15 (core PCE assumption 2.5% → neutral rate 4.20%; 2.0% → 3.735%) and computes 4.20-2.5=1.70 and 3.735-2.0=1.735 (0.035 percentage-point residual), showing that it is effectively of the form nominal neutral rate = real neutral rate + inflation assumption; Kiwoom Securities did not explicitly state that formula · judgment on August forecasts: monthly core PCE forecasts typically map detailed CPI and PPI components into PCE categories; because the mapping rule changes for the three components in Source 3, differences in institutional implementation may cause methodology effects to mix into forecast errors · this is the note's judgment, not the result of checking a specific institution's forecast; nor does it assume that major institutions continued using the old method unchanged after the BEA disclosed the revisions in June
14. Axios · Neil Irwin, Measurement tweaks will make inflation data look better 2026.07.06 · summarizes analysts' broad expectation as about a 0.2 percentage-point downward revision to core PCE ("consensus view that they will reduce core PCE inflation by about 0.2 percentage point") · calls the changes "entirely defensible on technical grounds" · notes that "the optics aren't great" when statistical-agency independence is under attack · this note does not use that wording as evidence of market pricing
15. Kiwoom Securities, "Reading the Fed Policy Redesign Through the Policy TF · A 5% Era for the U.S. 10-Year Treasury?" 2026.09.22 · the report's 4.20% nominal neutral-rate estimate adds a 2.5% core PCE assumption to the average of major models; with a 2.0% assumption, the estimate is 3.735% · 10-year estimates: 4.20+0.89=5.09% · 4.20+1.09=5.29% · stress 4.20+1.28=5.48% · term premium: Fed DKW model 0.78% at end-August · post-2000 long-run average 0.50% · +1 sigma 0.89% · report midpoint 1.09% · +2 sigma 1.28%.
16. Bank of Korea Economic Research Institute · BOK Economic Research 2026-15, "An Analysis of Korea-U.S. Long-Term Interest-Rate Comovement" posted 2026.09.20 (Yoon Jae-ho, Ewha Womans University · Kim Do-wan · Lee Hyung-seok, Bank of Korea) · official summary · finds global inflation shocks are the most important driver of comovement · external shocks are transmitted mainly through the expected future monetary-policy pathrather than risk compensation · suggests that appropriate expectations management can mitigate comovement to some extent · detailed contributions (global inflation 41.0% · U.S. long-term yields 22.7% · Fed monetary policy 18.3% · U.S. activity 18.0%) and the negative risk-compensation contribution for three of the four shocks were cross-checked against press reports from: NewsPim · Etoday · e-Focus
17. Federal Reserve Board · Chairman Kevin Warsh, Keynote remarks at the 2026 Jackson Hole Economic Policy Symposium 2026.08.28 · “Nor should we rely on isolated data points. Trends matter most.” · “we must interrogate reality to make sure we are not setting forward-looking policy based on stale or inaccurate data” · “I find it instructive to disaggregate the 199 individual components of the PCE price measure” · the speech does not directly discuss the BEA methodology change
2026.09.24 · Snowshagal INVESTMENT NOTE NO.05snowshagal.com