Which house price index is the best – MoneyWeek

Inanycase,giventheeurocrisis,weakgrowthintheUSandthedoublediprecessionintheUK,thereseemstobemoreimportantthingstoworryabout。

Thetroubleishousepricesarestilltoohigh。

MycolleaguePhilOakleyhasarguedthat,althoughproppedupbylowinterestrates,thehousingmarketisrottingfromwithin。

Anditlookasthough,beforethisyearisout,alleyesmaybeonhousingagain。

ThelatestdatafromNationwideclaimsthatpricesfellby0。

7%inJulyalone。

Moreimportantly,thiswasa2。

6%fallcomparedwiththesametimelastyear–thelargestannualdeclineseeninmorethanthreeyears。

Ofcourse,Nationwideisnttheonlycompanymeasuringhousepricestatistics。

Indeed,ourhouse-priceobsessednationproducesatleastsevendifferentindicespurelyonprices(thispiecedoesntdealwiththemonthlydatafromtheRoyalInstitutionofCharteredSurveyors,whichdoesntmeasurepricesdirectly)。

Thesepriceindicesareproducedby:Rightmove,Hometrack,OfficeforNationalStatistics,Nationwide,Halifax,LandRegistryandAcadametrics

Whicharethemostreliable?

Whicharethemostup-to-date?

Andarethereanyearlywarningsignalsthatcouldgiveyouaclueastowhenanothercrashisabouttobegin?

Letshavealook。

Askingprices:RightmoveandHometrack

TheRightmoveandHometrackindicesarebasedonaskingprices,thefirststageofthehomebuyingprocess。

Thesearethepricesthatthosehuntingforhomesonlinewillseefromasampleofestateagents(Hometrack)oronRightmoveswebsite(Rightmove。

co。

uk–whichcoversaboutathirdofthemarket)。

Intheory,thisshouldmaketheseindicesthetimeliestmeasures。

Theyshouldpickuppricetrendslongbeforetheyappearfurtherdowntheroadinfiguresbasedonmortgageapprovalsorcompletedsales。

However,therearealsosomeobviousdownsides。

Forastart,asyoucanseefromthelatestfigures(below),theaveragepricesproducedbyeachindexareverydifferent。

Presumablythisisdowntothesamplesize,butthegapisquitestriking。

Moregenerally,finalpricesareoftenverydifferenttotheinitialaskingprice。

Sellerscanhaveunrealisticviewsaboutthevalueoftheirproperty。

Weveallencounteredownerswhothinktheyshouldgetasmuchasthehousedowntheroad,evenalthoughitsinbetterconditionandhasalargergarden。

Also,iftheyboughtitforahigherpriceinthenearpast,itmaybedifficulttoacceptthattheywilllosemoneyonthedeal。

Estateagentssometimescontributetothis,usingunrealisticpriceestimatestogetpeopletosignupwiththem,onlytosuggestachangeafewweekslater(seePhilspieceonvaluinghousesformoreonthis)。

Thisdoesntjustcutonewayofcourse。

Ifapropertyisinanareaofhighdemand,thenitspossiblethatitmightsparkabiddingwar,drivingthefinalpriceup。

Andontopofallthis,theresnoguaranteethatjustbecauseahouseisonthemarketthatitllsell。

Theownermayhavetowithdrawit,thenputitonatalowerpriceatalaterdate。

Overall,accordingtoHometracksowndata,thecompletionpriceusuallyendsupbeing5-10%loweronaveragethantheaskingprice。

During2008,thefinaldiscountintheNorthofEnglandendedupbeingnearly15%。

Thesemeasures–thetimetakentosell,andtheaveragediscountfromoriginalprice–mightbemoreusefulthanaskingpricesforshowinghowmuchstressthemarketisunder。

Latestfigures

Rightmove:askingpricesfell0。

7%inJulyto£242,097,androse2。

3%onlastyear。

Hometrack:askingpricesfell0。

1%inJulyto£151,800,andfell0。

5%onlastyear。

Mortgageapprovalstage:ONS,NationwideandHalifax

TheOfficeforNationalStatistics(whotookoverfromtheDCLG),NationwideandHalifax(nowownedbyLloyds)allproducedatabasedonlendersvaluationsatmortgageapprovalstage。

Theideaisthatthisistimelierthandatabasedonthefinalsellingprice,butitisalsonearenoughtheendofthehome-sellingprocesstobeamorerealisticsnapshotthantheinitialaskingprices。

Theuseoflendersvaluationsalsomeansthatitisbasedonexpertopinion,filteringoutcaseswhereabuyer(orseller)wentwild。

However,someofthesestrengthscouldequallybeseenasflaws。

Likeallcompromises,collectingdatainthemiddleoftheprocess,meansthatitislesstimelythanaskingprices,butlessdefinitethanfinalsales。

Basingitonlendersvaluationsassumesthattheyareabetterjudgeofvaluethanbuyersandsellers。

Asweallknow,bankslentmoneylikedrunkensailorsupuntil2007,overvaluingmostproperties。

Manyhavearguedthattheyhavegonetoofarintheoppositedirectionsincethen,whichmeansthattheirvaluationshaveerredonthesideofcaution。

Wetakethepoint,althoughwedarguethatthebanksarerighttobemorecautious。

ThelargegapbetweentheONSfiguresandtheothertwoispartlydowntothedifferentmethodologyused–theONSfiguresarenotmix-adjusted,asweexplainbelow。

Latestfigures

ONS:housepriceswereunchangedinMayat£228,000,androse2。

3%onlastyear。

Nationwide:housepricesfell0。

7%inJulyto£164,389,andfell2。

6%onlastyear。

Halifax:housepriceswentup1%inJune,to£162,417andfell0。

5%onlastyear。

Completion:AcadametricsandLandRegistry

ThefinalapproachistobasethedataonagreedpricesfromtheLandRegistrysrecords(Acadametricsindexisbasedonthesamedatabutusesaslightlydifferentmethodology,whichproducesahigheraverageprice)。

Thisistheleastsubjectiveapproach,sinceitisntbasedonanestimatebyalender,orwhatabuyermightthinkthemarketcanbear。

Italsodoesntrelyonsampling(asinthecaseofHometrack),ordatafromonefirm(inthecaseofHalifax,NationwideandRightmove)。

However,thefactthatthepricesareonlytakenattheendoftheprocessmeansthattheunderlyingdataistheleasttimelyofalltheindices。

TheLandRegistryadmitsonitswebsitethatthereissometimesafurtherlagbetweenthesaletakingplaceandthedatabeingrecorded。

TheLandRegistrydataalsoincludescashsales,whichmaydistorttheirdata。

Thisisbecausetheyinvolvehigh-endproperties,whicharenotrepresentativeofthewidermarket。

BothindexescoverEnglandandWalesonly。

Latestfigures

Acadametrics/LSL:housepricesfell0。

1%inJuneto£224,102,androse3%onlastyear。

LandRegistry:housepricesrose0。

1%inJuneto£161,777,andfell0。

9%onlastyear。

Whatelseshouldyoubeawareofwhenusingthesefigures?

WiththeexceptionofHometrackandRightmove,alltheindicesareseasonallyadjusted,totakeaccountofthefactthathousepricesaregenerallyweakerinthewinterandstrongerinthesummer。

Similarly,sixofthesevenmeasures(theONSistheoddoneout)areadjustedtotakeaccountofthefactthathousesarenotidentical。

Totakeasillyexample:ifyousellahundredChelseatownhousesonemonth,thensellahundredterracesinNewcastlethenextmonth,itlllookasthoughpriceshavecollapsed。

Butallthatshappenedisyousoldaloadofveryexpensivehousesonemonth,andaloadofcheaponesthenextmonth。

Thattellsyounothingabouttheaveragehouseprice。

Sothefiguresaremix-adjustedtoaccountforthis。

Factorstakenintoaccountinclude(butarenotlimitedto);bedrooms,bathrooms,garden,location,parkingrights。

However,theONSindexisnotmix-adjusted。

Instead,itisasimpleaverageofthemortgagevaluationsinthemonth。

Theargumentisthattherearesomanytransactionsthatthedifferencesinthetypeofpropertysoldendupaveragingout。

Sowhichhousepriceindexisthebest?

Whatsreassuringisthatdespitethedifferencesinmethodologies,alltheindicestendtoproducesimilarresultsintermsofthetrendofprices,ifnotoftheactualpricesrecorded。

Asaresult,manyprefertousetheNationwidesindex,simplybecauseithasthelongesttrackrecord,havingstartedin1952。

However,intheshorterrun,thedifferencescanbeimportant。

Giventhattheyarebasedonaskingprices,itisnotsurprisingthattheRightmoveandHometrackmeasurestendtoleadtheothersbyuptosixmonths。

However,duringperiodsofvolatilitytheRightmoveandHometrackdatabecomemuchlessuseful。

Thisisbecausehouseholdstaketimetoadjusttheiraskingprices。

YoucouldalsoarguethattheONS,NationwideandHalifaxdatahavebecomelessusefulsincethebighousepricecrash。

Saleshaveplungeddrastically,andasaresulttheirsamplesizeshavefallentoo。