區(qū)域經(jīng)濟(jì)周期和房地產(chǎn)周期分析【外文翻譯】

1、<p>　　本科畢業(yè)設(shè)計(jì)（論文）</p><p>　　外 文 翻 譯</p><p><b>　　原文：</b></p><p>　　Regional Business Cycle and Real Estate Cycle Analysis and</p><p>　　The Role of Feder

2、al Governments in Regional Stability</p><p>　　For the last two decades the topic of the real estate cycle has gained a lot of attention not only in the fields of micro and macro economics, but also in the fi

3、eld of finance and investment. Recently real estate became a lucrative investment option for investors (Leonhardt [23]; Dhar and Goetzmann [15]). Securitization of the real estate market was one important trend that attr

4、acted many investors into this field. Further, now there are more investors who can participate in the global real esta</p><p>　　Many studies show that the real estate cycle has a direct impact on the behavi

5、or</p><p>　　of households, investors, banking systems, as well as on the national economy (Case [8], Wheelock [35], and Barlevy [1]). Very few studies, however, have compared and analyzed national and state

6、level business cycles with the national and regional real estate cycles. This comparison is important for at least three reasons: first, the clear idea about the national and state level real estate cycle will help home

7、owners and real estate investors minimize their losses. Second, it will help proper a</p><p>　　Switching estimation technique, this study compares the U.S. national and state level</p><p>　　busi

8、ness cycles with the U.S. national and state level real estate cycles. Second, depending on the formation of the state level real estate cycles, this study categorizes</p><p>　　different states, and _nally i

9、t analyses the severity of the state level real estate cycles.</p><p>　　The rest of the paper is organized as follows. First, we discuss related literatures, second we explain the data descriptions, third we

10、 provide model and methods, forth we give data description, fifth we state the results by presenting comparison of business cycles and real estate cycles, thus categorize states depending on the formation of real estate

11、cycles. To give some idea how the U.S. states’ real estate sector converges during the different phases of the real estate cycles, in section six</p><p>　　In the United States national business cycles are ca

12、lculated and dated by the National Bureau of Economic Research (NBER). Hamilton [20] used state space Markov Switching estimation technique on the U.S. GDP data to estimate business cycle turning points. Hamilton’s estim

13、ated dates coincided with the dates provided by the NBER which confirms the validity of the Markov Switching estimation technique to measure business cycle turning points. Bold in [3] compared with different business cyc

14、les turn</p><p>　　Crone [12], [13] used Kalman Filter estimation technique on the U.S. state level data and grouped U.S. into eight economic regions based on regional business cycles similarities. Using Hami

15、lton’s Markov Switching estimation technique on the state level coincident indexes6 Owyang, Piger andWall [27] and later Giannikos and Mona [16] dated the turning points of the U.S. state level business cycles. Both stud

16、ies show that the U.S. state level business cycles do not necessarily coincide with the nat</p><p>　　Exploring a threshold autoregressive (TAR) model Lizieri, Satchell, Worzala,</p><p>　　and Dac

17、co [24] found that regime switching model gives more accurate picture of real estate market performance than simple linear model. By using real interest rate as a state variable, they compare the behavior of the U.S. and

18、 the U.K real estate market. To measure the U.S. real estate market performance the authors used monthly data of the Real Estate Investment Trust (REIT) from December 1972 to March 1995. The U.K. real estate performance

19、was measured by the monthly data of International U.K</p><p>　　Proposing a simple model of lagged supply response to price changes and speculation in housing market Malpezzi and Wachter [25] generated real e

20、state cycles. They found that demand condition and speculation play major role in housing market</p><p>　　and real estate cycles. Further, they showed that the price elasticity of supply is the dominant comp

21、onent of speculation. The largest effects of speculation were observed when supply is inelastic.</p><p>　　Three different data sets are used in this study: 1) the U.S. fifty states coincident indexes; 2) the

22、 housing price indexes for the fifty U.S. states and the nation; and 3) national business cycle turning dates. Following are the descriptions of the data sets we used for this study.</p><p>　　The U.S. fifty

23、states’ monthly coincident indexes are provided by the Federal</p><p>　　Reserve Bank of Philadelphia dating from 1979:IQ . 2007:IIIQ. This data set is developed by Crone [12] estimating four latent dynamic f

24、actors of each state. The four variables are: the total number of jobs in nonagricultural establishments, average weekly hours in manufacturing, the unemployment rate, and the real wage and salary disbursements. This is

25、one of the most comprehensive monthly data set available for state level economic analysis. The reason for using state level coincident indexes f</p><p>　　The Housing Price Index (HPI) data used in this stud

26、y is published by the Office of Federal Housing Enterprise Oversight (OFHEO)8. The HPI is a broad measure of the movement of single-family house prices9, which measures weighted average changes in repeat sales, mortgage

27、defaults, prepayments, re_nancings, and housing affordability in specific geographic areas. The primary housing data are collected and provided by Fannie Mae and Freddie Mac to the OFHEO. The OFHEO generates HPI by using

28、 a modif</p><p>　　Housing Price Index (CQHPI)11. The HPI covers more transactions and geographic areas compared to other two data sets. We used quarterly HPI for fifty U.S. states from 1979:IQ to 2007:IIIQ.

29、</p><p>　　For the national real estate cycle analysis, we also used quarterly HPI data for the U.S. provided by the OFHEO. National business cycles and its turning points dates are listed by the NBER from 19

30、79:IQ to 2004:IQ. </p><p>　　First, we compare the national business cycle with the national real estate cycle. We use business cycle phases (e.g., recession and expansion) for our comparison. In the followin

31、g figures, vertical lines represent national recessions dated by the NBER. Recessionary states are measured in 0 to 1 scale, where 0 represents zero probability of recession, and 1 represents full probability of recessio

32、n. Therefore if the cycles are under 0.5 probability scale we called these the state expansion; those</p><p>　　Real estate recessions are marked by the solid (curve) lines in Figure 4.1. According to the Fig

33、ure 4.1, the U.S. has experienced two major real estate recessions during 1979:IQ to 2006:IQ period. One started at 1981:IIQ and ended at 1985:IQ, and the second one started at 1989:IVQ and sustained until 1999:IIIQ. For

34、 both cases, the real estate recessions started before the national recessions, and continued several periods after the national recession ended.</p><p>　　The result indicates that even though the real estat

35、e is one of the biggest industries in the United States, not all national recessions are due to the real estate sector fluctuation. In many cases real estate fluctuations may play an important role in some national reces

36、sions. Nevertheless, just from the Figure 4.1 alone, we cannot confirm that real estate was the sole reason of two national recessions of the1980s and the 1990s. Analyzing Figure 4.1, we also observe that in the recent y

37、ears, st</p><p>　　This paper also found that twenty two states15 have state level real estate cycle patterns similar to the national. In other words, these twenty two states experienced two real estate reces

38、sions as the nation during the 1980s and the 1990s; the span of recession, however, varied throughout the states. Nevertheless, the pattern of the cycle is not a sufficient condition for explaining the reasons behind sim

39、ilar real estate variables, other than real estates, are more responsible for the formation</p><p>　　The third group is a mixture of the leading and the lagging states. This set of states sometimes faces the

40、 leading real estate cycles, and sometimes faces the lagging real estate cycle compared to these states business cycles. Nine U.S. states19 fall into this category. In Figure 4.4: Maine, we observe that during the 1980s’

41、 and the 2000s’, economic and real estate fluctuations, the real estate cycle followed the state level business cycle with four and twenty quarters lags respectively. The 198</p><p><b>　　譯文：</b>&

42、lt;/p><p>　　區(qū)域經(jīng)濟(jì)周期和房地產(chǎn)周期分析</p><p>　　在過(guò)去二十年的房地產(chǎn)周期的話題，聯(lián)邦政府在地區(qū)穩(wěn)定中發(fā)揮的作用不僅獲得了在宏觀和微觀經(jīng)濟(jì)學(xué)領(lǐng)域，而且還得到金融和投資等領(lǐng)域的關(guān)注。最近房地產(chǎn)成為投資者有利可圖的投資選擇。房地產(chǎn)市場(chǎng)的證券化是一個(gè)重要的趨勢(shì)，吸引了許多投資者進(jìn)入這個(gè)領(lǐng)域。此外，現(xiàn)在比十年前更多的投資者參與全球房地產(chǎn)市場(chǎng)。然而，近幾十年來(lái)，世界經(jīng)歷了全球性房

43、地產(chǎn)，包括近期美國(guó)房地產(chǎn)危機(jī)波動(dòng)，這使得研究人員和投資者好奇房地產(chǎn)周期和結(jié)構(gòu)和他們是怎樣與國(guó)家相關(guān)以及遍布世界各地的其他經(jīng)濟(jì)活動(dòng)產(chǎn)生關(guān)聯(lián)。</p><p>　　許多研究表明，房地產(chǎn)對(duì)周期戶，投資者，銀行系統(tǒng)，以及對(duì)國(guó)家經(jīng)濟(jì)有直接影響。相關(guān)研究很少，但是，比較和分析了國(guó)家和區(qū)域房地產(chǎn)周期和國(guó)家和省級(jí)層面的商業(yè)周期，這個(gè)顯得比較重要，至少有三個(gè)原因：第一，關(guān)于民族和國(guó)家水準(zhǔn)的清晰的概念，房地產(chǎn)周期將幫助業(yè)主和房地產(chǎn)投

44、資者將損失減到最低。其次，這將有助于有關(guān)當(dāng)局（政府，抵押貸款經(jīng)紀(jì)人，銀行等）進(jìn)行有效的決策。三，未來(lái)研究人員將對(duì)各國(guó)的經(jīng)濟(jì)結(jié)構(gòu)產(chǎn)生生動(dòng)的了解和較好的房地產(chǎn)周期行為的理解。本文的重點(diǎn)，嚴(yán)格科學(xué)的房地產(chǎn)宏觀的角度，分析了房地產(chǎn)周期的模式。因此，這項(xiàng)研究有三個(gè)主要目標(biāo)。首先，利用馬可夫開(kāi)關(guān)估計(jì)技術(shù)，這項(xiàng)研究比較了美國(guó)的國(guó)家和省級(jí)層面與美國(guó)國(guó)家和州一級(jí)房地產(chǎn)景氣循環(huán)周期。二，關(guān)于國(guó)家一級(jí)房地產(chǎn)周期形成的不同，此研究歸類(lèi)不同的國(guó)家，nally分析

45、了國(guó)家一級(jí)房地產(chǎn)周期的嚴(yán)重性。</p><p>　　本文的其余部分組織如下。首先，我們討論相關(guān)文獻(xiàn)，第二我們解釋數(shù)據(jù)，第三，我們提供模型和方法，第四我們給出數(shù)據(jù)描述，第五根據(jù)我們國(guó)家的經(jīng)濟(jì)周期和房地產(chǎn)周期比較的結(jié)果，此分類(lèi)取決于國(guó)家的房地產(chǎn)形成周期。為了了解美國(guó)各州的房地產(chǎn)部門(mén)在房地產(chǎn)周期的不同階段如何匯集，第六段我們提供了一個(gè)收斂性分析，最后我們?cè)诘?節(jié)結(jié)束。</p><p>　　在美國(guó)

46、，全國(guó)商業(yè)周期，由國(guó)家經(jīng)濟(jì)研究局（NBER）日期計(jì)算。漢密爾頓[20]用美國(guó)GDP數(shù)據(jù)估計(jì)狀態(tài)空間馬爾可夫轉(zhuǎn)換技術(shù)，估計(jì)商業(yè)周期的轉(zhuǎn)折點(diǎn)。漢密爾頓的估計(jì)日期恰逢由國(guó)家經(jīng)濟(jì)研究局證實(shí)了馬可夫轉(zhuǎn)換估計(jì)技術(shù)的衡量商業(yè)周期的轉(zhuǎn)折點(diǎn)所提供的日期的有效性。大膽[3]比較不同經(jīng)濟(jì)周期結(jié)果變成美國(guó)測(cè)年法的轉(zhuǎn)折點(diǎn)。他總結(jié)說(shuō)，基于卡爾曼濾波算法和交換[20]估計(jì)技術(shù)漢密爾頓的馬爾可夫的股票和沃森的[20]，[20]實(shí)驗(yàn)商業(yè)周期指標(biāo)，優(yōu)于所有其他商業(yè)周期測(cè)年

47、法。</p><p>　　利用K這個(gè)人對(duì)美國(guó)國(guó)家數(shù)據(jù)的估算技術(shù)，CRONE根據(jù)美國(guó)地區(qū)經(jīng)濟(jì)周期的相似性將美國(guó)分成八個(gè)經(jīng)濟(jì)區(qū)。這兩項(xiàng)研究表明，美國(guó)州一級(jí)的商業(yè)周期并不一定配合國(guó)家商業(yè)周期。按科龍[14]最近的研究還估計(jì)，美國(guó)經(jīng)濟(jì)周期的擴(kuò)散指數(shù)基于國(guó)家級(jí)。他的研究結(jié)論是擴(kuò)散指數(shù)數(shù)據(jù)集，以更好地跟蹤預(yù)測(cè)區(qū)域或商業(yè)周期的轉(zhuǎn)折點(diǎn)。探索一門(mén)限自回歸（TAR）的模型。薩切爾和達(dá)科[24]發(fā)現(xiàn)，狀態(tài)轉(zhuǎn)換模型給出了更準(zhǔn)確的房地產(chǎn)市

48、場(chǎng)不是簡(jiǎn)單的線性模型的表現(xiàn)情況。通過(guò)使用狀態(tài)變量的實(shí)際利率，他們比較了美國(guó)和英國(guó)的房地產(chǎn)市場(chǎng)行為。為了衡量美國(guó)房地產(chǎn)市場(chǎng)表現(xiàn)，作者使用從1972年12月1995年3月的房地產(chǎn)投資信托基金（REIT）月度數(shù)據(jù)。英國(guó)房地產(chǎn)，從1975年1月至1995年8月的月度數(shù)據(jù)對(duì)國(guó)際英國(guó)房地產(chǎn)價(jià)格指數(shù)的性能進(jìn)行了測(cè)試。他們發(fā)現(xiàn)美國(guó)和英國(guó)的房地產(chǎn)制度的不同，因此，他們得出結(jié)論，實(shí)際利率發(fā)揮了作為房地產(chǎn)性能指標(biāo)在這兩個(gè)國(guó)家重要的作用，即在高利率樓價(jià)下跌急劇

49、制度，相反情況發(fā)生在較低的利率制度。同樣，卡諾和DeFina表明，利率變動(dòng)，貨幣當(dāng)局在整個(gè)地區(qū)有差別的統(tǒng)一對(duì)國(guó)家產(chǎn)生了影響。在建筑，房屋，基礎(chǔ)產(chǎn)業(yè)或房地產(chǎn)專(zhuān)業(yè)的地區(qū)相比，</p><p>　　對(duì)價(jià)格的反應(yīng)滯后供給的變化和住房市場(chǎng)投機(jī)，沃特Malpezzi提出了生成的房地產(chǎn)周期簡(jiǎn)單的模型。他們發(fā)現(xiàn)，在住房市場(chǎng)需求狀況和投機(jī)，房地產(chǎn)周期發(fā)揮了重要作用。此外，他們表現(xiàn)出的供給價(jià)格彈性是投機(jī)的主要組成部分。觀察到投機(jī)的最

50、大影響是供給缺乏彈性。三組不同的數(shù)據(jù)用于研究：1）美國(guó)五十個(gè)州同步指標(biāo); 2）對(duì)美國(guó)50個(gè)州和全國(guó)住房?jī)r(jià)格指數(shù)， 3）國(guó)家商業(yè)周期循環(huán)的日期。以下是我們?yōu)槊枋鲞@個(gè)研究所使用數(shù)據(jù)集。</p><p>　　美國(guó)五十個(gè)州每月的一致指數(shù)是由聯(lián)邦儲(chǔ)備銀行從1979:IQ . 2007:IIIQ。該數(shù)據(jù)集是由科龍發(fā)展[12]估計(jì)的每四個(gè)潛伏狀態(tài)的動(dòng)態(tài)因素。這四個(gè)變量有：非農(nóng)業(yè)就業(yè)總數(shù)的機(jī)構(gòu)，制造業(yè)平均每周工時(shí)，失業(yè)率，實(shí)際工

51、資和薪金支出。這是一個(gè)由國(guó)家級(jí)經(jīng)濟(jì)分析提供的最全面的月度數(shù)據(jù)。這項(xiàng)研究使用國(guó)家級(jí)一致指數(shù)的原因是，沒(méi)有月得州生產(chǎn)總值（GSP）的數(shù)據(jù)對(duì)全美國(guó)可用。普惠制數(shù)據(jù)是在每年的基礎(chǔ)上，但國(guó)家級(jí)經(jīng)濟(jì)衰退或擴(kuò)張可以在一年之內(nèi)開(kāi)始和結(jié)束。</p><p>　　在這項(xiàng)研究中所使用的房屋價(jià)格指數(shù)（HPI）數(shù)據(jù)是由聯(lián)邦住房企業(yè)監(jiān)督（OFHEO的）8辦公室公布的。對(duì)HPI是一個(gè)單戶住宅價(jià)格，這個(gè)價(jià)格加權(quán)平均變化重復(fù)銷(xiāo)售，抵押貸款違約，預(yù)

52、付款項(xiàng)并在特定地理區(qū)域的住房負(fù)擔(dān)能力的運(yùn)動(dòng)廣泛的衡量標(biāo)準(zhǔn)。主要的房屋數(shù)據(jù)的收集和由房利美和房地美提供給OFHEO的。HPI的涵蓋更多的交易和相對(duì)于其他兩個(gè)數(shù)據(jù)集的地理區(qū)域。對(duì)于全國(guó)房地產(chǎn)周期分析，我們還利用HPI的數(shù)據(jù)，每季度由美國(guó)OFHEO的提供。從1979年全國(guó)商業(yè)周期的轉(zhuǎn)折點(diǎn)和上市日期。</p><p>　　首先，我們比較了與全國(guó)房地產(chǎn)周期的國(guó)家的商業(yè)周期。我們比較經(jīng)濟(jì)周期階段（例如，經(jīng)濟(jì)衰退和擴(kuò)張）。在下

53、面的數(shù)字，垂直線代表由國(guó)家經(jīng)濟(jì)研究局日期國(guó)家經(jīng)濟(jì)衰退。經(jīng)濟(jì)衰退狀態(tài)測(cè)量0到1的規(guī)模，其中0表示經(jīng)濟(jì)衰退的可能性為零，1代表經(jīng)濟(jì)衰退帶來(lái)的可能性。因此，如果在0.5概率的周期，我們稱(chēng)為規(guī)模擴(kuò)張狀態(tài);那些規(guī)模0.5以上的可能性，我們稱(chēng)為國(guó)家的經(jīng)濟(jì)衰退。根據(jù)下面的數(shù)字，美國(guó)經(jīng)歷了四個(gè)主要的國(guó)家再1979年經(jīng)歷了經(jīng)濟(jì)衰退。兩次衰退在80年代初，第三個(gè)是在20世紀(jì)90年代開(kāi)始，最后一個(gè)在2000年初了。</p><p>　

54、　房地產(chǎn)衰退的標(biāo)志是固體（曲線），如圖4.1線。根據(jù)圖4.1，美國(guó)在1979年經(jīng)歷了兩個(gè)主要的房地產(chǎn)衰退：1979:IQ 到2006:IQ。一開(kāi)始在1981年IIQ和1985年結(jié)束：IQ,，而第二個(gè)在1989年開(kāi)始：IVQ并持續(xù)到1999年：IIIQ。對(duì)于這兩種情況下，房地產(chǎn)經(jīng)濟(jì)衰退開(kāi)始之前，國(guó)家經(jīng)濟(jì)衰退，并持續(xù)幾個(gè)階段后，國(guó)家經(jīng)濟(jì)衰退結(jié)束。</p><p>　　結(jié)果表明，盡管房地產(chǎn)是在美國(guó)最大的產(chǎn)業(yè)之一，但并非

55、所有國(guó)家的經(jīng)濟(jì)衰退是由于房地產(chǎn)行業(yè)的波動(dòng)。在許多情況下，房地產(chǎn)波動(dòng)可能在一些國(guó)家經(jīng)濟(jì)衰退的重要作用。不過(guò)，剛剛從圖4.1單，我們無(wú)法證實(shí)，房地產(chǎn)是兩個(gè)the1980s和90年代國(guó)家經(jīng)濟(jì)衰退的唯一原因。分析圖4.1中，我們也發(fā)現(xiàn)，在最近幾年，從2006年起：IIQ，又一個(gè)全國(guó)性的房地產(chǎn)衰退的可能性非常高。在房地產(chǎn)不景氣，2007年：IIIQ發(fā)生在國(guó)家一級(jí)和在大約四五美國(guó)五十州中（附錄4.A.2）國(guó)家的水平。但是，迄今只有19個(gè)州使國(guó)家級(jí)經(jīng)

56、濟(jì)進(jìn)入衰退的概率較高，但沒(méi)有國(guó)家經(jīng)濟(jì)衰退的跡象。</p><p>　　本文還發(fā)現(xiàn)，有二十二個(gè)國(guó)家級(jí)房地產(chǎn)周期模式類(lèi)似于國(guó)家。換句話說(shuō)，這二十作為兩個(gè)國(guó)家在經(jīng)歷了20世紀(jì)80年代和90年代兩個(gè)國(guó)家的房地產(chǎn)衰退，衰退的跨度，但是，各個(gè)國(guó)家不相同。然而，循環(huán)模式不是為了解釋變量，同類(lèi)房地產(chǎn)的背后，除了房地產(chǎn)以外的充分條件，更適合國(guó)家級(jí)經(jīng)濟(jì)周期的形成，這可能最終影響到這些國(guó)家房地產(chǎn)領(lǐng)域的投資?？偣灿辛鶄€(gè)州 - 阿拉巴馬州

57、，特拉華州，馬里蘭州，新墨西哥州，得克薩斯州和華盛頓 - 這個(gè)地區(qū)屬于此類(lèi)。在圖4.3：馬里蘭州，我們觀察了1981，1990年代和2007馬里蘭州的房地產(chǎn)周期，不僅遵循了國(guó)家級(jí)商業(yè)周期模式，它也遵循了國(guó)家具有滯后房地產(chǎn)周期的模式。在圖4.3的國(guó)家級(jí)經(jīng)濟(jì)周期的特點(diǎn)是純紅色line17（曲線）。在所有三種情況下，紅色實(shí)線后面國(guó)家和省級(jí)層面的房地產(chǎn)周期的標(biāo)志是綠色和藍(lán)色的虛線。第三集團(tuán)是一家領(lǐng)先和落后狀態(tài)的混合物。這組狀態(tài)有時(shí)面臨著領(lǐng)先的房