320.325 Inflaion and Business Cycles Chaper 14 Asse Marke 1
The mos imporan asse markes are he following: (i) Bond marke (ii) Sock marke (iii) Real esae marke (iv) Foreign exchange marke (nex lecure) 2
The Bond Marke: Yield curves Ineres rae Mauriy 3
4
5
6
7
Each Eurozone counry has is own yield curve. Ineres raes on Euro Area 10-year Governmen bonds are as follows: 12 Nov 11 5 Nov 12 26 Oc 13 26 Mar 14 Ausria 3.04 1.95 2.14 1.82 Belgium 4.37 2.41 2.61 2.15 France 3.19 2.22 2.33 2.09 Germany 1.73 1.44 1.77 1.57 Greece 27.9 18.22 8.81 6.86 Ialy 7.51 4.99 4.12 3.34 Neherlands 2.17 1.70 2.12 1.75 Spain 5.80 5.75 4.24 3.27 (Source: The Economis and he Financial Times) 8
Why do hese raes differ? i ITA = i GER + risk premium The risk on bonds comes from he following: (i) Inflaion/Exchange rae risk (ii) Defaul risk In he Eurozone he fear is ha highly indebed governmens will eiher defaul (ii) or leave he Eurozone, leading o inflaion and depreciaion of he currency (i). 9
Negaive Yields In 2012, negaive yields were observed on some German Bundesbank 6-monh bonds. This means ha he face value of he bond is less han he price a which i is purchased. How can his happen? I represens a fligh o safey. Invesors are worried he Eurozone will collapse, and ha i is herefore safer o hold German Bundesbank bonds han o have Euros siing in a bank deposi. 10
% per annum US Yield Curves Source: Burda and Wyplosz (2013) 11
The cenral bank ses he overnigh ineres rae (or cash rae) a which banks lend o each oher. I has only limied influence over longer-erm raes. The cenral bank can affec longer-erm ineres raes o some exen by inervening in he bond marke (eiher buying or selling longer-erm bonds). An invered yield curve is a siuaion where he yield curve becomes downward sloping as happened in he US in Jan 2007. 12
Wha Does I Mean When he Yield Curve Invers? An invered yield curve can arise when marke paricipans expec shor erm ineres raes o fall in he nex year or so. Remember: a fall in shor erm ineres raes implies a rise in shor mauriy bond prices. To see why his is so consider he case of an invesor wih a 2-year invesmen horizon. She can inves for wo years in a number of ways. (i) Buy a 1 year bond, When i maures buy anoher 1 year bond. (ii) Buy a 2 year bond. (iii) Buy a 3 year bond and sell i afer 2 years. 13
When shor mauriy bond prices are expeced o rise, sraegy (i) becomes less aracive and sraegy (iii) becomes more aracive. In euilibrium, he average invesor should be indifferen beween sraegies (i), (ii) and (iii). Hence invesors will need a higher reurn on sraegy (i) relaive o sraegy (iii) han before o make hem indifferen. This effec acs in he opposie direcion o he risk premium. If he expeced rise in shor-erm ineres raes is large enough, i can dominae he risk premium effec and cause he yield curve o become downward sloping. 14
Implicaions of an invered yield curve An invered yield curve is a leading indicaor of recession (usually abou 4 uarers laer). Why? Shor-erm ineres raes are procyclical. Cenral banks implemen expansionary moneary policy in a recession and conracionary moneary policy in a boom. Can yield curve inversions cause recessions? Perhaps. Banks make profis by riding he yield curve. They lend a higher ineres raes over long horizons (e.g., 25 year morgages) and borrow a lower ineres raes over shor horizons. 15
The profi comes from he differenial beween he long and shor-erm ineres raes. When he yield curve invers, his sraegy ceases o be profiable. Bank profis fall and hey may sop lending, which may creae a credi crunch. This in urn can cause invesmen o fall and send he economy ino recession. The banks sraegy of riding he yield curve also makes hem vulnerable o credi crunches. If liuidiy dries up, banks are no longer able o rollover heir shor-erm deb and may be forced o defaul. 16
Marke Efficiency Markes are efficien if i is no possible o obain profis by insananeously rading asses of eual risk. This is known as he no arbirage condiion 17
Sock prices Absracing from risk, he no arbirage condiion implies ha he yield on a sock should be eual o he yield on a bond. The yield on a sock is calculaed as follows: d r yield on safe bond dividend yield 1 capial gain oal reurn on shares no-profi condiion 18
19 1 1 1 1 1 1 1 d d r d r d r Rearranging his euilibrium condiion we obain he price of a sock:
20 There is a similar relaionship beween he price of shares in each of he nex periods and dividends and share prices in he subseuen periods: 1 1 2 1 2 3 2 1 1 1. d r d r d r ec
Recursively subsiuing ou he fuure sock prices, we ge an expression for he price of a sock in he curren period: 1 d 1 r 1 1 d 1 r 1r 1 2 2 1 d2 1 3 1 r 1 r 1 r 1 r ec. 1 i0 1 r i1 d i Price = he ne presen value of fuure dividend paymens 21
The formula is easily exended o allow for he greaer riskiness in holding socks by incorporaing a risk premium ino he sock reurn euaion: r yield risk on premium safe bond dividend yield which wih recursive subsiuion reduces o: d 1 capial gain oal reurn on shares 22
A higher risk premium (holding he expeced dividend sream fixed) herefore acs o reduce he marke price of asses. The euiy premium puzzle is he empirical resul ha he difference beween he long run average reurn on shares and bonds implies an implausibly large risk premium. 23
There is no euiy premium puzzle over he las 30 years due o he poor performance of euiies since he sar of he financial crisis. Over he period 1900-2005 he euiy premium was abou 3.5 percen per year. Based on sandard macro models, he level of risk aversion implied by his premium would reuire ha an invesor would be indifferen beween an eually likely payoff of $50,000 or $100,000 (wih expeced reurn of $75,000) and a cerain payoff of $51,200. (Mankiw and Zeldes, 1991) Such exreme risk aversion seems implausible. 24
Some socks are more risky han ohers. Example: small cap socks may be more risky han large cap socks for wo reasons. (i) Small cap socks are less liuid (ii) Small cap socks are more likely o go bankrup Hence small-cap socks should have a higher risk premium han large-cap socks. In he long run a porfolio of small-cap socks herefore should ouperform a porfolio of large-cap socks. This does no imply a violaion of he efficien markes hypohesis. 25
Diversificaion of risk: Risk can be reduced by consrucing a diversified porfolio. Example: Asse A Asse B Asse C Sae 1 100 100 100 Sae 2 100 0 50 Sae 3 0 100 50 Sae 4 0 0 0 By forming a porfolio consising of 50% asse A and 50% asse B we end up wih a porfolio wih payoffs as described by Asse C. C is less risky han A and B. Diversificaion makes no difference in saes 1 and 4, since A and B have he same payoffs in hese saes. I helps hough in saes 2 and 3. 26
The Subprime Crisis of 2007 Housing morgages in he US were colleced ogeher ino large pools, sliced up ino sandardized srips and sold o invesors. This process is referred o as securiizaion. Pooled and sliced morgages of his ype are also known as morgage backed securiies (MBS). Problems: (i) The morgage iniiaor sold he morgage on o a financial inermediary, who hen packaged hem ino MBS and sold hem o invesors. This creaed a principle-agen problem. 27
Bernanke raised his issue wih regard o he subprime morgage marke in May 2007, jus as he subprime crisis was beginning. The purchaser of an MBS is he principle. The morgage iniiaor is he agen. In he even of defaul i is he principle no he agen who bares he cos. The agen herefore does no care abou defaul risk. The exreme case was NINJA morgages (NINJA=no income, no job or asses). Buyers of MBS did no realize unil oo lae ha here was a principle-agen problem. 28
(ii) I was assumed ha he correlaion of defaul risks across individual morgages was uie low. Invesors failed o realize ha in he even of recession and falling house prices, many morgagees would sar defauling on heir morgages simulaneously (paricularly a he risky low end). (iii) The house was used as collaeral. If a morgagee defauled, hen he house would be sold and he credior would receive he proceeds. Hence i was assumed ha defauls would no really hur crediors. 29
When crediors sold foreclosed houses hey go less han hey expeced, since house prices were falling. These forced sales also aced o push down house prices even furher. Boom line: he risks of subprime morgages were much more correlaed han invesors realized. Pooling hem ogeher did lile o diversify risk. Also, he principle-agen problem arising from he decoupling of morgage iniiaion and bearing he risk of defaul mean ha each individual subprime morgage was more risky han invesors realized. Hence hese securiies ended up significanly over priced in he marke. 30
(iv) Even worse, noone was sure exacly how much overpriced hey were. This is because he securiies were so complicaed. MBS were ofen packaged ogeher wih oher deb producs like credi card deb and suden loans o form collaeralized deb obligaions (CDOs). The raing agencies Moody s, Sandard and Poor s and Fich gave he CDOs and MBS good credi raings unil 2007. No knowing he value of heir own balance shees, banks did no know how much hey could lend. Increased volailiy in he markes also increased he perceived level for risk, causing banks o furher cu back lending. 31
The resul was a severe credi crunch ha hreaened o drive no only insolven firms, bu any bu he mos liuid firms ino bankrupcy. Afer he collapse of Lehman brohers in Sepember 2008 even he paymen sysem wobbled. The whole financial sysem came close o collapse in lae 2008. MBS and he US Housing Bubble There was unil 2007 high demand for MBS securiies. 32
Invesors (including banks) saw MBS and CDOs as a new low risk asse class ha would allow hem o furher diversify heir porfolios. The high demand encouraged an increase in supply of MBS, hus eroding lending sandards. The sable macroeconomic environmen also encouraged marke paricipans o underesimae he level of sysemic risk. This effecively provided house buyers wih a huge amoun of addiional funding o buy houses, which helped rigger a housing boom. The Figure on he nex slide shows real house prices in he US from 1890 o 2010. 33
Source: Malkiel (2010), Bubbles in Asse Prices, CEPS Working Paper No. 200. 34
Speculaive bubbles: Assuming a fixed dividend d every period and ignoring he risk premium we have he fundamenal price of a sock: 1 i0 1 r i1 d d 1 1r 1r i0 2 3 d 1 1 1 1 1 r 1 r 1 r 1 r d 1 1 r 1 1 1 r i d 1 r 1 r 1 r1 d r 35 35
Suppose now ha he price of asses exceed he fundamenal value: d r Referring back o he original zero-profi condiion we have ha: r yield on safe bond d dividend yield 1 capial gain oal reurn on shares no-profi condiion r d d dr r 1 needs o be posiive o fill he gap 36 36
Thus any sock price higher han he fundamenal value, can be jusified by he expecaion of an even higher price of he sock in he nex period. This is an example of a self-fulfilling prophecy. The price of socks rise because hey are expeced o rise. This can lead o a speculaive bubble. In Jan 2013 Eugene Fama (a srong advocae of EMH) was asked he following uesion: Many people would argue ha he inefficiency [in he GFC] was primarily in he credi markes, no he sock marke ha here was a credi bubble ha inflaed and ulimaely burs. 37
Fama s reply is illuminaing. I don even know wha ha means. People who ge credi have o ge i from somewhere. Does a credi bubble mean ha people save oo much during ha period? I don know wha a credi bubble means. I don even know wha a bubble means. These words have become popular. I don hink hey have any meaning. My opinion: I would agree ha i can be hard o ell ha here is a bubble while i is happening, since he expeced earnings on an asse can also rise in a boom, pushing up he perceived fundamenal value. Once a bubble burss hough I would say i is uie clear i was a bubble. 38
Source: Blanchard, Macroeconomics, 4 h Ediion 39
Do Bubbles Invalidae he Assumpion ha Asse Markes Are Efficien? No necessarily. You canno know when a bubble will burs. Trying o profi from a bubble is risky since markes can say irraional longer han you can remain solven being agains a bubble. Also, here is a self-fulfilling prophecy elemen o asse markes. If you hink everyone else expecs he price o coninue rising hen you should expec his as well. 40