---
ticker: "KRHOUSECSI"
title: "South Korea House Price Outlook Consumer Survey Index"
unit: "index"
frequency: "Monthly"
source: "Katona (1951); Theil (1952)"
release: "Updated Monthly"
category: "Housing Market"
country: "KR"
language: "en"
canonical: "https://kred.dev/en/series/KRHOUSECSI"
license: "https://creativecommons.org/licenses/by-nc-nd/4.0/"
latest_value: 125.00
latest_date: "2026-08-01"
first_date: "2013-01-01"
observations_total: 164
observations_shown: 120
---

# South Korea House Price Outlook Consumer Survey Index

## Overview

A monthly read on how households expect house prices to move, showing which way housing sentiment leans.

## Key Figures

|  | Value | Date |
|---|---|---|
| Latest | 125.00 | 2026-08-01 |
| Change from previous | -2.00 | 2026-07-01 |
| Change over one year | +14.00 | 2025-08-01 |
| Highest on record | 132.00 | 2020-12-01 |
| Lowest on record | 61.00 | 2022-11-01 |
| Period covered | 2013-01-01 – 2026-08-01 |  |
| Observations | 164 |  |

## Recent observations

| Date | Value | Change |
|---|---|---|
| 2016-09-01 | 112.00 | +4.00 |
| 2016-10-01 | 114.00 | +2.00 |
| 2016-11-01 | 107.00 | -7.00 |
| 2016-12-01 | 97.00 | -10.00 |
| 2017-01-01 | 92.00 | -5.00 |
| 2017-02-01 | 92.00 | 0.00 |
| 2017-03-01 | 99.00 | +7.00 |
| 2017-04-01 | 103.00 | +4.00 |
| 2017-05-01 | 109.00 | +6.00 |
| 2017-06-01 | 116.00 | +7.00 |
| 2017-07-01 | 115.00 | -1.00 |
| 2017-08-01 | 99.00 | -16.00 |
| 2017-09-01 | 103.00 | +4.00 |
| 2017-10-01 | 110.00 | +7.00 |
| 2017-11-01 | 106.00 | -4.00 |
| 2017-12-01 | 106.00 | 0.00 |
| 2018-01-01 | 110.00 | +4.00 |
| 2018-02-01 | 112.00 | +2.00 |
| 2018-03-01 | 107.00 | -5.00 |
| 2018-04-01 | 101.00 | -6.00 |
| 2018-05-01 | 102.00 | +1.00 |
| 2018-06-01 | 98.00 | -4.00 |
| 2018-07-01 | 98.00 | 0.00 |
| 2018-08-01 | 109.00 | +11.00 |
| 2018-09-01 | 128.00 | +19.00 |
| 2018-10-01 | 114.00 | -14.00 |
| 2018-11-01 | 101.00 | -13.00 |
| 2018-12-01 | 95.00 | -6.00 |
| 2019-01-01 | 91.00 | -4.00 |
| 2019-02-01 | 84.00 | -7.00 |
| 2019-03-01 | 83.00 | -1.00 |
| 2019-04-01 | 87.00 | +4.00 |
| 2019-05-01 | 93.00 | +6.00 |
| 2019-06-01 | 97.00 | +4.00 |
| 2019-07-01 | 106.00 | +9.00 |
| 2019-08-01 | 107.00 | +1.00 |
| 2019-09-01 | 109.00 | +2.00 |
| 2019-10-01 | 115.00 | +6.00 |
| 2019-11-01 | 120.00 | +5.00 |
| 2019-12-01 | 125.00 | +5.00 |
| 2020-01-01 | 116.00 | -9.00 |
| 2020-02-01 | 112.00 | -4.00 |
| 2020-03-01 | 112.00 | 0.00 |
| 2020-04-01 | 96.00 | -16.00 |
| 2020-05-01 | 96.00 | 0.00 |
| 2020-06-01 | 112.00 | +16.00 |
| 2020-07-01 | 125.00 | +13.00 |
| 2020-08-01 | 125.00 | 0.00 |
| 2020-09-01 | 117.00 | -8.00 |
| 2020-10-01 | 122.00 | +5.00 |
| 2020-11-01 | 130.00 | +8.00 |
| 2020-12-01 | 132.00 | +2.00 |
| 2021-01-01 | 130.00 | -2.00 |
| 2021-02-01 | 129.00 | -1.00 |
| 2021-03-01 | 124.00 | -5.00 |
| 2021-04-01 | 122.00 | -2.00 |
| 2021-05-01 | 124.00 | +2.00 |
| 2021-06-01 | 127.00 | +3.00 |
| 2021-07-01 | 129.00 | +2.00 |
| 2021-08-01 | 129.00 | 0.00 |
| 2021-09-01 | 128.00 | -1.00 |
| 2021-10-01 | 125.00 | -3.00 |
| 2021-11-01 | 116.00 | -9.00 |
| 2021-12-01 | 107.00 | -9.00 |
| 2022-01-01 | 100.00 | -7.00 |
| 2022-02-01 | 97.00 | -3.00 |
| 2022-03-01 | 104.00 | +7.00 |
| 2022-04-01 | 114.00 | +10.00 |
| 2022-05-01 | 111.00 | -3.00 |
| 2022-06-01 | 98.00 | -13.00 |
| 2022-07-01 | 82.00 | -16.00 |
| 2022-08-01 | 76.00 | -6.00 |
| 2022-09-01 | 67.00 | -9.00 |
| 2022-10-01 | 64.00 | -3.00 |
| 2022-11-01 | 61.00 | -3.00 |
| 2022-12-01 | 62.00 | +1.00 |
| 2023-01-01 | 68.00 | +6.00 |
| 2023-02-01 | 71.00 | +3.00 |
| 2023-03-01 | 80.00 | +9.00 |
| 2023-04-01 | 87.00 | +7.00 |
| 2023-05-01 | 92.00 | +5.00 |
| 2023-06-01 | 100.00 | +8.00 |
| 2023-07-01 | 102.00 | +2.00 |
| 2023-08-01 | 107.00 | +5.00 |
| 2023-09-01 | 110.00 | +3.00 |
| 2023-10-01 | 108.00 | -2.00 |
| 2023-11-01 | 102.00 | -6.00 |
| 2023-12-01 | 93.00 | -9.00 |
| 2024-01-01 | 92.00 | -1.00 |
| 2024-02-01 | 92.00 | 0.00 |
| 2024-03-01 | 95.00 | +3.00 |
| 2024-04-01 | 101.00 | +6.00 |
| 2024-05-01 | 101.00 | 0.00 |
| 2024-06-01 | 108.00 | +7.00 |
| 2024-07-01 | 115.00 | +7.00 |
| 2024-08-01 | 118.00 | +3.00 |
| 2024-09-01 | 119.00 | +1.00 |
| 2024-10-01 | 116.00 | -3.00 |
| 2024-11-01 | 109.00 | -7.00 |
| 2024-12-01 | 103.00 | -6.00 |
| 2025-01-01 | 101.00 | -2.00 |
| 2025-02-01 | 99.00 | -2.00 |
| 2025-03-01 | 105.00 | +6.00 |
| 2025-04-01 | 108.00 | +3.00 |
| 2025-05-01 | 111.00 | +3.00 |
| 2025-06-01 | 120.00 | +9.00 |
| 2025-07-01 | 109.00 | -11.00 |
| 2025-08-01 | 111.00 | +2.00 |
| 2025-09-01 | 112.00 | +1.00 |
| 2025-10-01 | 122.00 | +10.00 |
| 2025-11-01 | 119.00 | -3.00 |
| 2025-12-01 | 121.00 | +2.00 |
| 2026-01-01 | 124.00 | +3.00 |
| 2026-02-01 | 108.00 | -16.00 |
| 2026-03-01 | 96.00 | -12.00 |
| 2026-04-01 | 104.00 | +8.00 |
| 2026-05-01 | 112.00 | +8.00 |
| 2026-06-01 | 120.00 | +8.00 |
| 2026-07-01 | 127.00 | +7.00 |
| 2026-08-01 | 125.00 | -2.00 |

## Definition

KRHOUSECSI is a raw series recorded without transformation, holding as reported the forward-looking component of a qualitative survey that asks households which way house prices will move. The index is a net-response, diffusion-index statistic that subtracts the share of respondents expecting declines from the share expecting increases and then standardizes the result about a neutral benchmark of 100, so that a reading above that mark means households expecting increases outnumber those expecting declines.

The qualitative design that admits only up, same, or down answers, and the theoretical link that the balance of such replies bears to realized outcomes, was formalized by Anderson (1952). The balance statistic that condenses trichotomous replies into a single continuous reading is an established quantification (Theil 1952), later systematized through probability thresholds and critically reviewed (Carlson and Parkin 1975; Nardo 2003).

The premise that household attitudes and expectations are themselves observable and economically consequential determinants of spending is the starting point of such consumer survey design (Katona 1951). Treating the resulting net-response figure as a tool of cyclical measurement follows the diffusion-index tradition (Moore 1961), operationalized for monthly real-time signaling (Shiskin 1961).

The reference cycle against which such a series is read was defined within the program of cyclical-indicator measurement (Mitchell and Burns 1938; Burns and Mitchell 1946), and monthly series of this kind were later recast as the latent common factor of comoving aggregates (Stock and Watson 1989; Stock and Watson 1991).

The series therefore belongs to the family of expectational attitudinal gauges, whose construction and meaning have been reviewed across the literature (Zarnowitz 1992; Ludvigson 2004; Koopmans 1947).

## Methodology

KRED applies no transformation to KRHOUSECSI, neither rescaling, deflation, seasonal adjustment, smoothing, nor annualizing, and stores the published monthly figure exactly as received. The methodology to describe is therefore limited to the survey-and-index basis by which such a figure comes to exist.

The starting point is a qualitative tendency survey in which households answer with an ordinal up-same-down judgment rather than reporting the size of the price change they expect. The theoretical link between the balance of replies and underlying conditions was set out by Anderson (1952), and the premise that attitudes and expectations can be treated as observable data was established by Katona (1951).

The balance, namely the share answering up minus the share answering down, summarizes these ordinal responses in a single number (Theil 1952). The probability thresholds under which trichotomous answers map to a quantitative series were formalized, and the competing balance, probability, and regression quantifications were catalogued (Carlson and Parkin 1975; Nardo 2003). The resulting statistic is presented on the diffusion-index scale about a neutral value, so that its sign and level carry the monthly signal (Moore 1961; Shiskin 1961).

The indicator-selection logic that justifies treating an expectational survey as a cyclical gauge derives from the cyclical-indicator program, and reproducible turning-point dating was established as an algorithmic procedure (Mitchell and Burns 1938; Burns and Mitchell 1946; Bry and Boschan 1971). The dynamic-factor formalization of the common cyclical signal and the general design principles of household surveys are likewise in place (Stock and Watson 1991; Hussmanns, Mehran, and Verma 1990).

KRED performs no rescaling, deflating, smoothing, annualizing, or seasonal adjustment, so the recorded monthly figure is the survey index itself. The standing caution that such atheoretical measurement is not a structural model remains (Koopmans 1947).

## Applications in Economics

The house price outlook index is a timely summary of how households see prices ahead, and it carries expectations that move before housing demand and the borrowing decisions attached to it. The premise that measured attitudes and expectations help drive spending is what supports this use (Katona 1951).

Its standing as a cyclical gauge derives from the program of cyclical-indicator measurement (Mitchell and Burns 1938; Burns and Mitchell 1946), a framing evaluated across the coincident, leading, and diffusion families and operationalized for real-time signal detection (Zarnowitz 1992; Shiskin 1961; Moore 1961).

The cyclical signal embedded in such a series is reinterpreted as the latent common factor of comoving monthly aggregates (Stock and Watson 1989; Stock and Watson 1991), and its turning points are identified by reproducible procedures (Bry and Boschan 1971).

Empirically, survey sentiment carries information about household spending beyond what hard data already explain (Carroll, Fuhrer, and Wilcox 1994; Ludvigson 2004). Because housing expectations form alongside perceptions of employment and income, the series is read next to labor aggregates resting on the operational definitions of the labor force and the flow view of unemployment (Hussmanns, Mehran, and Verma 1990; Clark and Summers 1979).

The series describes expectations rather than forecasting a price path, and it must be read with the measurement caveats attaching to a quantified balance of qualitative replies (Theil 1952; Carlson and Parkin 1975; Nardo 2003; Koopmans 1947).

## Applications in Financial Markets

For market participants the house price outlook index is high-frequency sentiment data that arrives ahead of transaction and price statistics, so its level relative to the neutral benchmark and its turns are watched as early signals of housing-related demand.

The turning-point logic that gives such a series its forward content was proposed in the cyclical-indicator program and made algorithmic (Mitchell and Burns 1938; Burns and Mitchell 1946; Bry and Boschan 1971), and monthly diffusion readings have been treated as real-time signals (Shiskin 1961; Moore 1961; Zarnowitz 1992).

Whether a qualitative sentiment reading carries predictive content for spending beyond hard data has been tested and surveyed, resting on the behavioral premise that expectations measurably shape demand (Carroll, Fuhrer, and Wilcox 1994; Ludvigson 2004; Katona 1951).

Household expectations are not market prices, so the figure must be read with the properties of a balance statistic (Theil 1952; Carlson and Parkin 1975; Nardo 2003) and interpreted within the single-factor representation of comoving activity (Stock and Watson 1991; Stock and Watson 1989).

The series is a descriptive sentiment gauge rather than a tradable forecast, and the caution against mistaking a measured indicator for a structural model applies (Anderson 1952; Koopmans 1947).

## Statistical Tests

KRHOUSECSI is a not-seasonally-adjusted monthly house price outlook survey index, tested over 162 observations from 2013-01 to 2026-06 under a constant specification without trend, since the level fluctuates around a nonzero mean with no deterministic trend. The survey scale is bounded but never binds over the sample, so the standard and bounded-support unit-root nulls nearly coincide and no bounded correction is needed (Cavaliere and Xu 2014).

The augmented Dickey and Fuller (1979) test with the lag augmentation of Said and Dickey (1984) rejects a unit root at p = 0.0118, the nonparametric Phillips and Perron (1988) test also rejects at p = 0.035, and the KPSS test, whose null is stationarity, does not reject with a statistic of 0.145 at p > 0.10 (Kwiatkowski et al. 1992), so all three procedures concur on a clean I(0) classification. The DF-GLS escalation of Elliott, Rothenberg, and Stock (1996) is reserved for ambiguous readings and is not required on this clean verdict.

On the clean I(0) verdict the level is the stationary object, so its mean-break dates are citable. The Bai and Perron (1998) procedure, computed by the algorithm of Bai and Perron (2003), locates four mean shifts, at 2015-12, 2019-10, 2022-06, and 2024-06, read as level movements of a stationary series rather than as stochastic-trend breaks (Perron 1989). On the stationary level the Ljung and Box (1978) portmanteau, refining the original Box and Pierce (1970) form, returns Q = 396.9 at lag 12 and Q = 555.297 at lag 24 and rejects the no-autocorrelation null at p = 0.000 at both rungs, confirming the strong serial dependence expected of a persistent expectations level.

The index is not seasonally adjusted, and the seasonal battery is the designed diagnostic. The HEGY test rejects seasonal unit roots with a statistic of 154.897 at p = 0.000 (Hylleberg et al. 1990; Beaulieu and Miron 1993), so the seasonal pattern present is not stochastic. The seasonal-dummy F test on the differenced series, however, returns 1.842 at p = 0.0518 and therefore does not reject at the 5% level, while the QS statistic at seasonal lags 12 and 24 of the same differenced series rejects with 69.393 at p = 0.000, so seasonal dependence remains that fixed monthly means do not summarize. Only the verdict that the seasonality is not stochastic is settled, and its characterization as deterministic dummies is provisional, a contrast handled within the stationary-seasonality null and the seasonal-time-series framework (Canova and Hansen 1995; Ghysels and Osborn 2001).

## Frequently Asked Questions

### Which housing market indicators are published for Korea?

Dwelling sale prices, leasehold deposit prices, the stock of unsold dwellings, housing transaction counts and the housing sentiment survey, each as a published monthly level.

### Why does the stock of unsold new housing matter?

It is inventory completed but not cleared, so it accumulates when demand falls short of supply. Prices are downward sticky while inventory responds first, which makes it a signal that runs ahead of price adjustment.

### How does the jeonse price index differ from the sale price index?

The sale price is the price of ownership and the leasehold deposit is the price of deposit-based occupancy. Their ratio expresses the relative cost of holding versus using housing, and it can stay divergent for long periods.