Housing Forecasts via Stock Market Indicators Varun Mittalaand Laura P. Schaposnikb Corresponding author schaposuic.edu

2025-05-06 0 0 1.4MB 9 页 10玖币

侵权投诉

Housing Forecasts via Stock Market Indicators

Varun Mittalaand Laura P. Schaposnik?,b,

(?) Corresponding author: schapos@uic.edu

Through the reinterpretation of housing data as candlesticks, we extend Nature Scientiﬁc Reports

article by Liang and Unwin [LU22] on stock market indicators for COVID-19 data, and utilize some

of the most prominent technical indicators from the stock market to estimate future changes in

the housing market, comparing the ﬁndings to those one would obtain from studying real estate

ETF’s. By providing an analysis of MACD, RSI, and Candlestick indicators (Bullish Engulﬁng,

Bearish Engulﬁng, Hanging Man, and Hammer), we exhibit their statistical signiﬁcance in making

predictions for USA data sets (using Zillow Housing data) and also consider their applications within

three diﬀerent scenarios: a stable housing market, a volatile housing market, and a saturated market.

In particular, we show that bearish indicators have a much higher statistical signiﬁcance then bullish

indicators, and we further illustrate how in less stable or more populated countries, bearish trends

are only slightly more statistically present compared to bullish trends.

Keywords: Stock market indicators, housing market, trend study

I. INTRODUCTION

The mathematical study of housing markets dates back

to the 1970s [Kir76], and its prominence has seen a sig-

niﬁcant increase in the last decades, as consumers are

increasingly looking at the right time to make a signiﬁ-

cant investment [GLB19]. Whilst classical studies have

included complex models and economic examinations to

varying levels of success [Car89], after the housing market

crashed in 2008 it became clear that new models are nec-

essary to understand both the local and global behaviour

of housing markets. Since then, more novel approaches

begun to be implemented [Kha08], but these are yet to

become as robust or successful as they could potentially

be [KM71]. Purchasing a house is never easy, and buying

at the optimal time is a signiﬁcant decision. As a result

of COVID, the housing market has become more volatile

and has ﬂuctuated signiﬁcantly [LS21]. Therefore, we

propose the utilization of technical indicators that are

commonly applied in the stock market to the housing

markets to identify both bearish and bullish trends in

the housing market, where

(a) a trend is considered bullish if it is corresponds to

an increase in value for a time series.

(b) a trend is considered bearish if it corresponds to

a decrease in value for a time series.

In a recent Nature Scientiﬁc Reports article, Liang and

Unwin proposed to use stock market indicators to under-

stand trends in COVID-19 cases in the US [LU22]. In-

spired by this work, we shall consider Zillow’s data set of

the USA’s housing market and study it through examin-

ing the usability of statistical indicators usually seen in

the stock market such as MACD Bearish, MACD Bullish,

RSI Bearish, RSI Bullish, Hanging Man, and Hammer.

These indicators can be used to identify trends (both

bearish and bullish) that can be vital when it comes to

house purchasing decisions. In what follows we shall show

via these indicators, trends (both bearish and bullish) can

be detected and used to understand housing prices.

For each indicator we calculate the statistical signif-

icance using the Wilcoxon test, and show in particular

that while both RSI indicators were statistically signiﬁ-

cant, only the bearish indicator is signiﬁcant for MACD.

Finally, we extend our analysis to other countries, with

the intent of being able to make insights in various types

of markets. To be encompassing and robust, we consid-

ered diﬀerent types of markets (Saturated, Volatile and

Stable) and studied the relevance of the indicators within

each setting. In particular, we show that

•across all countries, bearish indicators were much

more statistically signiﬁcant compared to bullish

indicators, for the vast majority of indicators..

•the diﬀerence in the statistical signiﬁcance of

bullish and bearish trends isn’t nearly as large in

saturated and volatile housing markets.

•when using sock market indicators directly on hous-

ing prices, the trends of the housing price can be

predicted more accuratedly than when considering

the indicators on housing ETFs.

To carry out our study, we considered RSI, MACD,

and Candlestick Analysis conducted through R. After

grouping grouping Zillow’s data [zil21] into Heikin Ashi

candlesticks, we scanned the data for the four candle-

stick indicators mentioned above and used the Wilcox,

and later performed a similar procedure for the other in-

dicators. After introducing some background in Section

II, we dedicate Section Vand Section VI to the main

ﬁndings of our work, which can be seen in two diﬀerent

directions in terms of the indicators themselves and their

statistical signiﬁcance in various markets. To highlight

the utility of our approach vs any standard stock mar-

ket study of Housing ETFs, we compared our ﬁndings

to what one can deduce through the same indicators on

the most used Housing ETFs, namely VNQ (Vanguard

Real Estate ETF), SCHH (Schwab US REIT ETF), and

XLRE (Real Estate Select Sector SPDR Fund) in Section

??. Finally, we expand on the analysis and applications

of the above ﬁndings in Section VII. Finally, we also at-

tached our code for the project on Github [MS22].

arXiv:2210.10146v1 [econ.GN] 18 Oct 2022

II. BACKGROUND: STOCK MARKET

INDICATORS

Researchers have considered a large number of ways

that stock market prices can be forecasted, for example

by speculating a stock’s intrinsic value, employing com-

plex algorithms, and speculating human behavior [CP00],

or by using some complex algorithms such as rolling win-

dow analysis and deep learning [MST19]. Moreover, re-

cently there has been some success around more simple

but nuanced indicators, popularly known as candlestick

analysis, MACD, and RSI [CNL14]. Stock indicators are

being used increasingly to identify trends in the stock

market, as traders are looking to maximize their proﬁts

by trying to gain insight on the future of the market.

The current research is inspired by the work in [LU22]

where stock market indicators were used to forecast

COVID-19 cases, and thus we shall dedicate this sec-

tion to review some of the main results which shall prove

useful for our research and for comparison with our own

ﬁndings. The indicators considered in [LU22] were can-

dlestick analysis, RSI and MACD indicators. After giv-

ing a mathematical deﬁnition of those indicators for a

time series of datapoints, the authors measured their sta-

tistical signiﬁcance by ﬁrst considering the null hypoth-

esis, the probability that an external circumstance was

the cause of a bearish or bullish trend, not an indicator

(recall that an indicator is described as bullish if that

indicator predicts a future decrease while an indicator is

described as bearish if it describes a future increase in

prices). This probability was denoted as a p-value, and

it was then ascertained that lower p-values translated to

higher statistical signiﬁcances as that meant the prob-

ability of a trend being forecasted by an indicator was

higher in comparison to a random external event. After

considering this hypothesis, they used the Wilcoxon test

for R to determine these p-values across all of the indica-

tors. This allowed them to make conclusions about the

applicability of indicators.

In what follows we will give a brief overview of the

work in [LU22] on candlestick analysis, RSI and MACD

indicators introduced for the timed datasets of COVID-

19 cases, and we will then extend to the timed housing

data of [zil21].

A. Candlestick Analysis

The ﬁrst type of indicators considered were candlestick

indicators, which included Hammer, Bullish Engulﬁng,

Dark Cloud Over, Bearish Engulﬁng, and Hanging Man.

FIG. 1: Candlestick patterns studied in [LU22].

Weekly candlesticks were constructed with the real

body constituting the region marked by the endpoints

of the opening and closing values (covid cases), and the

upper and lower limits being the maximum and mini-

mum cases over the time interval respectively. Then, the

authors developed R code was developed to identify the

candlestick patterns shown in Figure 1 within the dataset

to identify the frequencies and predictive capabilities of

each of these patterns occurred.

B. RSI

The RSI (Relative Strength Index), is a popular indi-

cator that is similar to considering the recent “momen-

tum” of a time series through average rate of increases

and decreases in a market, in this case the market in con-

sideration (housing, COVID, or stock). The indicator is

constructed by dividing the closing values Ctover some

period into two sets:

•The set Gtin which the series increased:

Gt=Ct−Ct−1

, Ct> Ct−1.

•The set Dtin which the series decreased:

Dt=Ct−1−Ct

Ct< Ct−1.

From the above sets one can compute the averages Atand

Btusing the EMA Vnover nperiods with a smoothing

factor of 1

nleading to

At=Vn[Gt] and Dt=Vn[Dt].

Then, the RSItindicator at time tis deﬁned as follows

RSIt= 100 −100

1 + G0

Within the stock market, the indicator becomes useful

when the RSI value exceeds a lower and upper threshold,

as that is when the stock in consideration becomes over-

bought or oversold, as seen in Figure 2. If the RSI value

falls below the lower threshold, that signiﬁes a stock be-

ing oversold, which signals a bullish trend as stock prices

will soon increase in price. Conversely, if the RSI value

exceeds an upper threshold, that signiﬁes a stock being

overbought, signaling a bearish trend as stock prices will

fall.

FIG. 2: RSI indicator on Zillow Stock.

文档加载中……请稍候！
如果长时间未打开，您也可以点击刷新试试。

下载文档到电脑，查找使用更方便

10 玖币 0人已下载

立即下载

摘要：

HousingForecastsviaStockMarketIndicatorsVarunMittalaandLauraP.Schaposnik?;b,(?)Correspondingauthor:schapos@uic.eduThroughthereinterpretationofhousingdataascandlesticks,weextendNatureScienticReportsarticlebyLiangandUnwin[LU22]onstockmarketindicatorsforCOVID-19data,andutilizesomeofthemostprominenttec...

展开>> 收起<<

Housing Forecasts via Stock Market Indicators Varun Mittalaand Laura P. Schaposnikb Corresponding author schaposuic.edu.pdf

共9页,预览2页

还剩页未读，继续阅读

声明：本站为文档C2C交易模式，即用户上传的文档直接被用户下载，本站只是中间服务平台，本站所有文档下载所得的收益归上传人(含作者)所有。玖贝云文库仅提供信息存储空间，仅对用户上传内容的表现方式做保护处理，对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私，请立即通知玖贝云文库，我们立即给予删除！

Housing Forecasts via Stock Market Indicators Varun Mittalaand Laura P. Schaposnikb Corresponding author schaposuic.edu

相关推荐

开通VIP享超值会员特权

作者详情

相关内容

热门标签

举报选择: