The impact of equity analyst recommendations on market attention, price-consensus and the behaviour of other analysts

Lotter, Rousseau (2015-12)

Thesis (PhD)--Stellenbosch University, 2015.

Thesis

ENGLISH ABSTRACT: Analysts are valuation specialists who advise both institutional clients and non-professional investors on the choice and timing of security purchases and sales. The analysts’ advice may have hugely beneficial or unfavourable outcomes for those who rely on them. This study investigated the possible influence of 901 local and international analysts’ recommendations that were issued from 1993 to 2011 on shares listed on the Johannesburg Stock Exchange (JSE). The short-term impact of recommendations on prices and possible behavioural tendencies among analysts, including a reported inclination to issue overly-positive recommendations, were respectively investigated in the first two empirical chapters. Thirdly, the success rate of analysts to issue recommendations with an advised directional impact and possible herding behaviour among analysts were researched. The empirical chapters conclude with an investigation into changes in investor attention (as proxied by traded volumes) and price volatility around analysts’ recommendations. The efficient market hypothesis and the ‘differences of opinion’ theories were used as fundamental points of departure and interpretation. More than 37 000 recommendations, ranging from strong buy to strong sell, were used in an event-study methodology to analyse the market’s reaction to these recommendations. Advanced modelling techniques were implemented in Excel and VBA to analyse daily consensus opinions, positive- versus negative sentiment, analyst activity and reactions, the frequency of abnormal price reactions, abnormal price movements, abnormal traded volumes, and changes in price volatility surrounding recommendation revisions. The study found that analyst recommendations were followed by an abnormal reaction in prices and that the magnitude of a recommendation’s change (e.g. a three-step change from strong sell to buy versus a one-step change hold to buy) had a greater impact than a recommendation’s absolute level. A portfolio strategy revealed the possible benefit of recommendations for investors. Analysts issued their opinions using different patterns within the five possible recommendation categories, and issued the same proportion of negative recommendations during periods of low business confidence and economic contraction than during growth- and economic upswing phases. Analysts who issued more recommendations in total were not more influential than less active analysts, and not all analysts were able to issue recommendations with a large advised directional abnormal impact. As expected, recommendations that had a large abnormal price impact generated some herding activity among the other analysts who covered the same share. Investor attention increased around the issuance of recommendation revisions, and price volatility increased after large recommendation upgrades. In support of market efficiency, investors seemed able to trade at new price levels and execute their trades with sufficient liquidity following recommendations. Results that infer differences of opinion were present both among analysts and investors: competing analysts did not issue the same recommendations for the same shares and favoured different recommendations categories; and investors only acted on some of the recommendations. Furthermore, analysts did not have the same propensity to cause abnormal price reactions. Traded volumes increased around recommendation revisions, showing that investors paid attention to recommendations.

AFRIKAANSE OPSOMMING: Analiste spesialiseer in die waardasie van maatskappye en adviseer beide institusionele- en nie-professionele beleggers rakende die keuse en tydsberekening van hul kope en verkope. Díé advies kan baie voordelige of nadelige gevolge hê vir diegene wat daarop staatmaak. Hierdie studie het die moontlike invloed ondersoek van 901 Suid-Afrikaanse en internasionale analiste se aanbevelings rakende JSE-genoteerde aandele tussen 1993 en 2011. Die eerste twee empiriese hoofstukke ondersoek (i) die korttermyn impak van analiste se aanbevelings op pryse en (ii) moontlike gedragspatrone onder analiste, insluitend ‘n gerapporteerde neiging om oor-positiewe aanbevelings uit te reik. Derdens is analiste se sukseskoers om aanbevelings met ‘n verwagte impak uit te reik en moontlike ‘trop’-gedrag onder analiste nagevors. Die empiriese hoofstukke sluit af met ‘n ontleding van veranderinge in beleggers se aandag (soos aangedui deur verhandelde volumes) en prysvolatiliteit rondom analiste se aanbevelings. Die effektiewe markhipotese en die ‘verskil in opinie’ teorie was gebruik as fundamentele grondslag en om resultate te interpreteer. ‘n Gebeurtenis-studie metodologie is gebruik om die mark se reaksie op meer as 37 000 aanbevelings, wat van sterk koop tot sterk verkoop strek, te analiseer. Gevorderde modelleringstegnieke is in Excel en VBA geïmplementeer om konsensus opinies, positiewe- vs. negatiewe sentimentsperiodes, analiste se aktiwiteitsvlakke en reaksies, abnormale prysreaksies en die voorkoms daarvan, abnormale verhandelde volumes, en veranderinge in prysvolatiliteit rondom aanbevelings hersienings te bereken en te analiseer. Die studie het bevind dat analiste se aanbevelings wel gevolg is deur abnormale prysbewegings, en dat die grootte van aanbevelings se hersienings (bv. ‘n drie-stap hersiening van sterk verkoop na koop versus ‘n een-stap hersiening van hou na koop) ‘n groter impak as die aanbeveling se absolute vlak gehad het. ‘n Portefeulje strategie het ook die moontlike voordeel van aanbevelings vir beleggers uitgelig. Analiste het verskillende patrone binne die vyf-punt aanbevelingskategorieë gebruik om hul opinies te kommunikeer, en het dieselfde proporsie negatiewe aanbevelings tydens periodes van swak besigheidsvertroue en ekonomiese afswaai uitgereik as tydens periodes van groei en ekonomiese opswaai. Analiste wat meer aanbevelings in totaal uitgereik het, was nie meer invloedryk as ander analiste nie, en nie alle analiste het aanbevelings wat ‘n groot abnormale prysreaksie veroorsaak het, uitgereik nie. Soos verwag het aanbevelings, wat groot abnormale prysbewegings veroorsaak het (invloedryke aanbevelings), ‘trop’-gedrag veroorsaak onder kompeterende analiste. Beleggers se aandag het toegeneem met die uitreik van hersienings, en prysvolatitliteit het toegeneem ná groot aanbeveling-opgraderings. Beleggers kon teen nuwe prysvlakke verhandel en hul besluite uitvoer met genoeg likiditeit nadat aanbevelings uitgereik is, wat indikatief van mark-effektiwiteit is. Resultate dui ook op verskillende opinies tussen beleggers en analiste: analiste het verskillende aanbevelings vir dieselfde aandele uitgereik en het verskillende aanbevelings-kategorieë verkies, en beleggers het nie op alle analiste se aanbevelings gereageer nie soos aangedui deur pryse en volumes. Analiste het verder nie dieselfde geneigdheid gehad om abnormale prysveranderinge te veroorsaak nie. Verhandelde volumes het toegeneem rondom aanbevelingshersienings, wat aandui dat beleggers wel aandag aan die analiste se aanbevelings gegee het.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/97986
This item appears in the following collections: