Asia Market Analysis
Why is household spending weak when wages are improving?
A falling benchmark alongside broad gains, and improved wages alongside lower household spending. Aligning populations and time periods clarifies the demand picture in Japan and wider Asia.
October 9, 2026 edition | Information cutoff: October 9, 2026, 15:34 JST; after the Tokyo cash close
1．Tokyo markets: reading a falling index alongside widespread gains
Swipe horizontally to compare the items.
- Major benchmarks
- Market breadth
- Income and expenditure
The Nikkei 225 closed on October 9 at 69,030.92, down 11.19 points, or 0.02%. TOPIX rose 13.35 points, or 0.33%, to 4,104.81. These are final values for the Tokyo cash session ending at 15:30 JST, not futures prices or intraday observations. The Nikkei recovered most of its 529-point morning loss but still finished marginally lower. TSE Prime recorded 1,193 advancers, 328 decliners and 25 unchanged issues, with 25 of 33 sectors rising. Marine transport, other products, services and food led the sector gains; banks, nonferrous metals and retail led the declines. A small benchmark loss thus coexisted with broad gains across stocks and sectors. [26][27][28][29]
The starting point for understanding Tokyo on October 9 is the divergence between a major benchmark and the breadth of individual share-price moves. The Nikkei 225 ended the morning at 68,512.30, down 529.81 points, or 0.77%. Yet the TSE Prime market had 1,218 advancers, 302 decliners and 26 unchanged issues. Kabutan's 11:31 a.m. report showed gains in 21 of 33 sectors. A substantial benchmark decline coexisted with gains across many companies during the morning session. [12]
JPX's 12:01 p.m. update showed TOPIX at 4,092.50 after the morning session, up 1.04 points, or 0.03%. TOPIX had edged higher while the Nikkei fell. [26] Divergence between the two does not imply that either display is wrong. Their constituents and the weights translating each company's price movement into the benchmark differ. The same session can therefore produce different aggregate pictures.
The Nikkei is a price-weighted index of 225 companies, with price-adjustment factors and a divisor maintaining continuity. TOPIX is weighted by free-float market capitalization. Free float takes account of the shares potentially available for trading. Even a company included in both benchmarks need not have the same influence in each. Differences between the indices help identify where changes are concentrated. [9][10]
The number of advancing stocks measures breadth by giving each company one vote. A small gain in a small company counts as one advancer, just as a large gain in a large company does. Broad participation in gains therefore does not establish that aggregate market capitalization rose. Conversely, a few influential declines can pull an index down while prices rise across many other business areas. Benchmarks, advance-decline counts and sector distributions measure the same market on different scales.
Bloomberg's 9:42 a.m. market report attributed pressure on semiconductor-related shares and other areas to concerns about AI profitability following reports concerning OpenAI's revenue. That is the news organization's assessment of a factor attracting market attention. It does not mean that this report independently verified the private company's detailed revenue or contracts, or that actual orders across related industries fell simultaneously that day. Revisions to expectations and subsequently observable business activity operate on different timelines. [11]
Equity prices reflect expectations of future profits as well as current revenue, and the interest rates used to translate future earnings into present values. Where expectations of growth were high, their revision can produce conspicuous price movements. Equipment, customer usage and actual payments, however, follow contracts and implementation processes. Prices can move first without factory utilization or employment changing by the same proportion on the same day. This timing gap is fundamental to reading the real economy through financial markets.
The same morning report placed Japan's ten-year government bond yield at 3.035%, down 4.5 basis points, with the yen trading in the low 158s per dollar. One basis point equals 0.01 percentage point. These were morning observations, not final daily values. Falling bond yields alongside a falling equity benchmark illustrate the limits of explaining share-price direction through interest rates alone. [11]
At 15:00 JST, dollar/yen was around ¥158.16, approximately ¥0.02 lower than at 17:00 the previous day and therefore almost unchanged. This is a time-specific observation reported by Minkabu Press at 15:24, not a closing value for the foreign-exchange market as a whole. A cash-equity close and continuously traded currencies require different comparison clocks. [30]
Bond prices rise when yields fall, but the interest paid by a company to its bank does not move one-for-one with that day's government bond yield. Renewal dates, fixed or floating terms, credit conditions and loan maturity also matter. Exchange rates raise a similar timing issue. This morning's yen rate and an import contract priced months earlier have different implications for cash payments. Market updates describe current conditions; existing contracts take time to transmit them.
Today's domestic data shift attention to what households spent and what those payments purchased. August's household survey showed real consumption expenditure below its year-earlier level. Previously released wage data showed improvement, but total household income and earnings per employee measure different populations. In both equities and household finances, aligning the unit of measurement reveals concrete structures behind apparent contradictions. [1][2]
Tokyo's cash sessions run from 9:00 to 11:30 a.m. and from 12:30 to 3:30 p.m. Morning breadth can change in the afternoon, while overseas markets may still be trading when Tokyo closes. The market and timestamp behind each figure determine its meaning. Presenting the day's close alongside the morning's developments shows how assessments evolved within the session. [8]
2．Household spending: why an annual decline can coexist with a monthly increase
| Indicator | Period and value | Population and comparison |
|---|---|---|
| Consumption expenditure | August: ¥310,975 [1] | Average per household with two or more members |
| Real consumption expenditure | August: −3.1% year on year [1] | Seasonally adjusted monthly change: +0.1% |
| Worker-household actual income | August: ¥616,704 [1] | Nominal +1.3% year on year; real −0.9%, excluding imputed rent |
| Cash earnings per employee | August: ¥311,364 [2] | Establishments with 5+ employees; nominal +3.8% year on year |
| Real wages | August: +1.5% year on year [2] | Deflated by all-items CPI excluding owners’ equivalent rent |
| Economy Watchers | September: current 47.0; outlook 47.4 [18] | Seasonally adjusted; monthly changes +0.6 and −0.9 point |
Swipe horizontally to compare the items.
- Nominal payments
- Price-adjusted expenditure
- Comparison periods
The Statistics Bureau's August Family Income and Expenditure Survey, released on October 9, put average consumption expenditure by households with two or more people at ¥310,975. Nominal expenditure fell 1.0% from a year earlier, while real expenditure, adjusted for prices, fell 3.1%. Seasonally adjusted real spending rose 0.1% from the previous month. The annual comparison was weak; the monthly comparison showed an increase close to flat. Both can hold because the reference periods differ. [1]
Year-on-year comparisons match the same season in successive years. Monthly comparisons capture the latest change, normally after adjusting for recurring seasonal patterns. A high year-earlier base can leave annual growth negative despite small recent monthly improvements. Conversely, a depressed base can produce positive annual growth even when the current recovery is weak. Using both comparisons helps distinguish the pace of improvement from the level already reached.
Nominal spending records the money paid. Real spending adjusts for price changes to approximate movements in the quantity of goods and services purchased. When prices rise, an unchanged payment may buy fewer items or less service. The difference between the nominal and real declines demonstrates why payments alone do not capture changes in purchasing power. Aggregate real spending nevertheless remains an estimate using price indices, rather than a direct count of every product purchased.
The ¥310,975 figure is an average for households with at least two members. It is neither expenditure per Japanese resident nor total national consumption including single-person households. Household size, age, housing tenure and the composition of earners affect spending. Applying the national average directly to an individual family's living costs introduces differences between the survey population and that family's circumstances. [17]
The survey's breakdown shows real declines across several categories of everyday goods and services. The Statistics Bureau reported decreases in food, housing, furniture and household utensils, and culture and recreation, among other categories, while transport and communication increased. Broad categories themselves contain both rising and falling expenditures. The aggregate decline cannot establish whether households postponed replacements, used fewer services or shifted payment dates. [17]
Durables are not purchased every month, so monthly averages can be sensitive to how many sampled households made a purchase. Postponing a car or appliance until the following month reduces current expenditure without necessarily eliminating the underlying need. Recurring payments for food, transport and utilities behave differently. Understanding a monthly change requires viewing recurring expenditure and large purchases in light of their respective frequencies.
Seller-side statistics have a different scope. Retail sales are business-side totals; the household survey measures household-side averages. Retail figures include visitor purchases and some business sales, while household spending also includes services and transfers between households. Higher shop sales need not imply an equivalent increase in expenditure by the average resident household. This distinction affects conditions in tourist destinations differently from areas dominated by residents' everyday purchases. [17]
Among worker households with two or more members, average actual income was ¥616,704, up 1.3% in nominal terms. It fell 0.9% when deflated by the all-items CPI excluding owners' equivalent rent, and 0.6% using the all-items CPI. The survey's term “actual income” does not itself mean inflation-adjusted income. Nominal actual income and its separately calculated real growth rates are reported alongside one another. [1]
Owners' equivalent rent values the housing services enjoyed by homeowners by reference to comparable rent. It captures housing services even where no cash rent is paid, but its relevance differs when examining cash outgoings. The choice of price index changes measured real income growth. Switching deflators midway through a comparison makes it difficult to identify whether income changed or the calculation basis changed.
The relationship between wages and household income extends the October 7 issue's discussion of base-related pay, overtime and bonuses to the number of earners in each family. One person's pay can rise while another member's hours fall, producing a different movement in total household receipts. Considering headcount and income-source composition allows a more coherent comparison between the household survey and wage releases. [21]
Each purchasing-power measure has a particular coverage. Wage statistics capture payments to employees, the household survey captures money received and spent by families, and sales statistics capture revenue received by businesses. They observe the same economy at different points and need not move identically. The latest household survey specifically shows that expenditure on the household side remained below its year-earlier level.
For illustration, suppose last year's purchases cost 100 and the same basket now costs 102. Spending 100 again would buy the equivalent of about 98.0 of last year's quantity, calculated by dividing 100 by 1.02. Nominal spending can be unchanged while real spending falls. Spending 103 instead would buy roughly 101.0 in real terms. This hypothetical example explains nominal and real measures; it is not a reconstruction of the survey. Official calculations use category-specific prices and expenditure composition, so applying one headline inflation rate to total spending need not reproduce the published result.
3．From higher pay to spending: earners, obligations and confidence
Swipe horizontally to compare the items.
- Earnings per employee
- Household disposable income
- Recurring purchases
The preliminary Monthly Labour Survey released on October 7 put August total cash earnings per employee at ¥311,364 in establishments with at least five employees, up 3.8% year on year. Real total earnings, deflated using the CPI excluding owners' equivalent rent, rose 1.5%. This can coexist with falling real income among worker households because the populations differ. Pay from an establishment to an employee and the income received by all members of a household differ in headcount and non-wage income. [2]
The establishment survey aggregates employee earnings. A household may contain employees, self-employed people, pensioners and students. Even a household classified as a worker household need not consist entirely of people with identical employment arrangements. Earnings of the household head and other members, bonus timing and changes in the number of earners affect the total. An average wage figure alone cannot determine total household receipts.
Disposable income is the money available after taxes and social insurance contributions, among other deductions. Assessing funds available for choice requires moving from gross pay to this stage. Even higher disposable income need not all become immediate consumption. Debt repayment, provision for future large payments and rebuilding cash balances offer several uses for the same increase in income.
An increase in regular pay and a one-off receipt create different confidence about the future. Income viewed as persistent can make recurring commitments or durable purchases easier to undertake. A temporary receipt may instead clear an overdue need or rebuild reserves. Consumption can respond later than wage data because households assess whether the improvement will last.
Slower inflation is also different from a return to earlier price levels. Accumulated price increases can leave living costs high even after the current inflation rate declines. An initial pay increase may first fill gaps left by past cost increases rather than immediately restore every purchase previously cut. This adjustment is one possible reason for weak expenditure even in a month when annual real-wage growth improves.
Household spending baskets differ from the average. Families with high food, utility and commuting costs face a different experience from those with high housing or education bills. The people receiving higher wages need not be the people facing the strongest price increases. Evaluating household pressure therefore requires attention to expenditure composition and the recipients of income gains, alongside average earnings growth.
Businesses also face a lag between pay increases and selling-price adjustments. A company raising wages to retain staff may experience pressure on margins and cash flow if it cannot immediately raise prices. Raising prices can, in turn, reduce quantities purchased. Higher income for employees, higher costs for employers and changed prices for consumers are three sides of the same development. Wage increases consequently feed through the whole income-expenditure cycle.
Productivity improvement can ease this constraint. Producing more value with the same workforce and hours creates room to raise pay while containing unit costs. Buying equipment alone does not deliver the complete result. Work processes, utilization, training, maintenance and available orders determine the outcome. Investment spending measures the input; actual output per hour and shorter lead times measure what it achieves.
In its October 8 Regional Economic Report, the Bank of Japan assessed all regions as recovering or picking up. Investment and employment-income conditions were improving in some areas, while housing investment remained weak. A nationwide spending statistic and regional interview-based assessments can create different impressions because their populations, periods and criteria differ. Strong areas of activity can coexist with weak aggregate household expenditure. [3]
September's Economy Watchers Survey showed the seasonally adjusted current-conditions DI rising 0.6 point to 47.0, while the outlook DI fell 0.9 point to 47.4. Current conditions are assessed against three months earlier; the outlook concerns the next two to three months. More respondents can perceive current improvement while becoming more cautious about the future. Such differences in confidence also affect when additional income is spent. [18]
The previous issue's external income surplus concerned the structure of Japan's receipts from overseas. Today's household survey moves the focus to expenditure by domestic families. [4] Corporate earnings, wages, disposable income and consumption form a connected sequence, but recipients and uses change at each stage. Identifying which stages improve and which lag provides a more useful explanation than reducing the economy to a single strong-or-weak label. [20]
Average earnings per person and the aggregate wage bill can also move differently. In a hypothetical fixed population, a 3% increase in average pay alongside a 2% decline in employee numbers would raise the wage bill by approximately 0.94%, since 1.03 multiplied by 0.98 equals 1.0094. Lower headcount offsets part of the average improvement. More employees can instead expand the total even with modest average gains. This is not a claim that Japanese employment fell 2% in the month; it demonstrates why headcount is needed to translate an average into aggregate income.
4．Asian trading calendars and Chinese demand: different clocks on the same date
Swipe horizontally to compare the items.
- China: second reopened session
- Taiwan and Korea: closed
- Time zones and demand end uses
Mainland Chinese equity markets resumed normal trading on October 8 after the October 1–7 National Day closure. October 9 is the second session after reopening. The Shanghai Stock Exchange's calendar identifies Saturday, October 10 as a weekend closure. Administrative make-up working days in China do not necessarily correspond to exchange trading days. Reading the calendar market by market is essential to assigning prices to the correct date. [5]
The Taiwan Stock Exchange observes October 9 as an adjusted holiday because National Day, October 10, falls on a Saturday. October 9 is Hangeul Day in Korea, covered by the Korea Exchange's public-holiday closure rules. Neither cash equity market has a new October 9 closing price. A number remaining on a screen may refer to the latest completed session or another instrument, rather than a daily change comparable with Tokyo's October 9 close. [6][24][25]
Corporate news, currencies and related assets listed abroad can move during a local exchange closure. The local cash market incorporates that information after reopening. Similarly, US semiconductor-share movements need not already be reflected in that day's local cash indices when assessing Taiwan's and Korea's semiconductor industries. Industrial connections and trading continuity differ. An exchange closure neither suspends all economic activity nor removes exposure to overseas developments.
Mainland China and Hong Kong are one hour behind Japan. Tokyo's 3:30 p.m. cash close is 2:30 p.m. in China and Hong Kong. This time difference, alongside each market's trading arrangements, affects daily comparisons. Chinese and Hong Kong prices may still be intraday observations when Japan's confirmed close is available. Later overseas price movements cannot retrospectively explain a Tokyo session that had already ended.
China's September manufacturing PMI, published by the National Bureau of Statistics, was 50.1, up 0.3 point from August and above 50. The PMI aggregates purchasing managers' responses; 50 is a reference point for the balance of reported improvement and deterioration. A reading of 50.1 does not mean production grew 0.1%. It uses a different unit from growth in actual output or sales. [13]
The large-company PMI was 50.6, compared with 49.7 for medium-sized firms and 48.9 for small firms. The production index was 51.7 and new orders were 50.5. An aggregate reading above 50 does not establish uniform improvement across business sizes. When larger firms occupy different positions in supply chains and financing, smaller businesses can face different order books and cash conditions despite an improving headline. [13]
The survey put the main raw-material purchase-price index at 60.8 and the output-price index at 54.0. These measures capture the breadth of reported price changes, not a 60.8% rise in input costs. Nor can their difference be treated as a corporate profit margin: input and output composition and respondent conditions differ. They provide an entry point into cost and selling-price trends, while the resulting effect on profits requires other evidence. [13]
Rising Chinese factory production can transmit orders to Asian suppliers of components and machinery. But greater production can reflect inventory replenishment as well as final customer demand. Concentrated shipping dates around holidays can also make one week's figures appear strong. Combining production, orders, inventories and exports helps establish why activity increased and whether it can persist.
For Japanese businesses, different uses matter even within trade with China. Consumer goods, factory equipment and components for export processing have different ultimate customers. Weak Chinese household spending can coexist with orders supported by production for overseas markets. Conversely, a domestic services recovery need not immediately generate imported-equipment demand. What is sold, and to whom, explains the commercial relationship more precisely than a general statement that China is strong.
Australia, India and Southeast Asia face both shared costs and distinct domestic demand. Energy exporters and importers, and foreign-currency borrowers and earners, can be affected in opposite directions by the same commodity or dollar movement. Compressing the region into one equity-index direction obscures these differences between receipts and payments. The broader the geographic comparison, the more useful trade structure and financing currency become.
On October 7, the International Energy Agency announced member support for accelerating the stock releases agreed in March and prioritizing diesel where possible. It said completing all previously pledged but unreleased volumes would bring approximately 100 million barrels to market. This was not a new decision to add another release of that size. Asian importers care about when required fuel products actually arrive, as well as the aggregate amount of crude available. [23]
Fuel reaches users through ships, ports, refineries and storage facilities. Stocks still need transport when release locations and demand centers are apart. The emphasis on diesel reflects product-specific conditions that differ from total crude supply. Energy links Japanese living costs, Chinese manufacturing expenses, and transport costs in India and Southeast Asia, making it a common foundation for understanding regional markets.
Converting export receipts into local currency can also introduce changes unrelated to volume. If foreign-currency revenue is unchanged while the local currency depreciates, reported local-currency sales rise. If imports become more expensive in the same currency, the effect on profit can be offset. Comparing Asian nominal export values requires identifying currency translation, unit prices and physical quantities. These components connect balance-of-payments and trade data to the actual workload of factories.
5．Preparing for the next releases: levels, changes and policy lags
Swipe horizontally to compare the items.
- October 12: Tokyo closed
- October 14: US CPI
- October 29–30: BOJ meeting
This morning's household survey describes August spending. September's Economy Watchers Survey, October's Regional Economic Report and October 9 equity prices refer to different periods. Ordering information by publication date does not recreate the order in which economic events occurred. Statistics measure past activity with a lag; markets incorporate future expectations into current prices. Retaining reference periods prevents one weak release from being treated as a complete description of the present.
A recovery in household expenditure can take several forms: more units purchased, delayed large purchases resumed, or spending shifted toward services. Each can increase the total, but the implications for business procurement and staffing differ. Broad quantity growth is more likely to affect inventories and hours worked, while a few large purchases can create volatile monthly results. Returning from the aggregate to its components helps identify the economic activity likely to follow.
Payments can also rise while quantities fall. Higher prices may increase the amount spent even as consumers reduce items or frequency. A company watching revenue and a service provider counting visits may therefore perceive different conditions. A larger basket per shopping trip alongside fewer customers or visits differs from sustained demand expansion. Combining prices, quantities and frequency makes revenue more informative.
Statistical revisions also matter. Preliminary Monthly Labour Survey figures can change in final releases, while seasonal adjustment and rebasing can revise historical data. CPI rebasing also affects real growth rates in household expenditure. Comparing consistent series rather than mixing old published values with revised observations helps distinguish economic changes from statistical ones. A persistent trend is more informative than a fine difference in decimal places. [1][2]
The next monetary policy publication date is an opportunity for additional information, not a predetermined conclusion. The Bank of Japan schedules its next policy meeting for October 29–30, with the Outlook Report's basic view due on October 30. Household spending, wages, business pricing and overseas costs enter the overall assessment. No mechanical rule translates one August consumption release into the meeting's decision. [14]
The US September CPI release is scheduled for October 14 at 8:30 a.m. Eastern Time, or 9:30 p.m. in Japan under daylight saving time. Information released after Tokyo's cash close can first affect currencies and futures before reaching cash equities in the next session. Treating daytime and overnight releases as simultaneous obscures what each price could have incorporated. [19]
In his October 8 speech, Federal Reserve Governor Christopher Waller anticipated additional rate increases if incoming data developed as expected, while retaining flexibility over their pace. His example of three 25-basis-point increases illustrated central-bank communication. Interpreting it as a commitment to a fixed sequence changes its meaning. Decisions, personal forecasts and illustrative assumptions have materially different implications when reading policy statements. [15]
The October 9 Europe & the Americas issue examined how interest rates and energy costs reach business payments. Overseas costs also enter Japanese household conditions through imports, logistics, hiring and selling-price adjustments. The purpose of reading foreign indicators is to understand changes in the terms on which Japan buys, sells and finances activity, rather than assuming identical economic outcomes across countries. [22]
October 12 is Sports Day in Japan, and Tokyo's cash market is closed. Not all overseas markets close simultaneously, so information continues to accumulate while domestic cash prices remain unchanged. Cash-market holidays and holiday trading in derivatives follow different arrangements. Checking the instrument, exchange and reference date prevents a prior-session price displayed as the latest figure from being mistaken for that day's result. [7]
Businesses also operate with deadlines different from news cycles. Counterparty holidays, bank-processing dates and shipping cutoffs follow their own operational clocks. An exchange holiday does not suspend every activity in a country, but it helps frame the timing of funding and price checks. Connecting economic commentary with business operations requires attention to settlement conditions as well as trading conditions.
Today’s Market Takeaways
Today's market takeaway is that benchmark direction does not reveal the breadth of corporate price moves, and average wages do not fully describe household purchasing power. Annual spending declines, a small monthly increase, improved wages and regional recovery assessments can coexist when their populations and periods are aligned. Preserving what was measured and when, including Japanese and other Asian trading calendars, is the foundation for understanding current conditions.
Analysis & Commentary
The rest of this article is paid content
Read only this article
Unlock the paid section of this article with a one-time purchase.
One-time payment Pay Per Article l Daily Asia Market Analysis
$1.00
Payment is completed on Stripe's secure checkout page.
Subscriptions renew automatically and can be canceled anytime. Cancellation terms are described in the Legal Notice (Commercial Transactions).
Verify your access without leaving this page.
This browser does not support passkeys. Please use another method.
Subscriptions and purchased articles
Check your subscription, purchased articles and payment method without leaving this page.
JavaScript is required to manage subscriptions and purchases here.
Subscriptions and purchased articles
Review and manage your Reader Account here. You stay on this article the whole time.
Sign in to your Reader Account to see your subscriptions and purchased articles.