‘The Dis­crete Charm of Se­cu­ri­ti­za­tion’: A Die-Hard Story

Dünya Executive - - ANALYSIS - Gunduz FINDIKCIOG­LU Chief Econ­o­mist

Can banks se­cu­ri­tize NPLs? It was ages ago that the idea of se­cu­ri­ti­za­tion was first con­ceived in the minds of fund man­agers and trust hold­ers. How­ever, in its mod­ern guise the pack­ag­ing of a first CDO can be traced back to 1983, although the idea of a CDO is a bit older, i.e. 1980. It is con­tem­po­ra­ne­ous with the ad­vent of Rea­gan and Thatcher, and what has been dubbed ne­olib­er­al­ism in many cir­cles. It changed the whole struc­ture of Western Cap­i­tal­ism, and ended the so­cial com­pact of Roo­sevelt. It caused an in­cred­i­ble in­crease in wealth and in­come dis­tri­bu­tions – although CEO pay­checks in­flated the wage bill and made it ap­pear less dra­mat­i­cally twisted than it ac­tu­ally was - and winded the wings of fi­nan­cial­iza­tion to such a dra­matic scale that de-in­dus­tri­al­iza­tion be­came the hall­mark for at least two decades - true or not, that is. Nev­er­the­less, one thing is cer­tain: ser­vices grew by leaps and bounds, tra­di­tional in­dus­tries shrank and fi­nan­cial mar­kets be­came al­most to­tally un­fet­tered de­spite the SEC and all that.

The­ory…

Now, here is the sit­u­a­tion: In a model with lin­ear real as­sets, bud­get con­straints are ho­mo­ge­neous in prices ( jointly in as­set and com­mod­ity prices) so prices can be de­fined up to a nor­mal­iza­tion. This is the usual Ar­row-De­breu story. Fur­ther­more, as in the con­text of Ar­row-De­breu reg­u­lar economies, we ob­tain generic lo­cal unique­ness re­sults. In the case of nom­i­nal as­sets, how­ever, there is a con­tin­uum of equi­lib­rium al­lo­ca­tions. In the ab­sence of a well-de­fined “mon­e­tary sys­tem” such in­de­ter­mi­nacy may oc­cur. One im­pli­ca­tion is that the in­de­ter­mi­nacy of equi­lib­ria trans­lates into that of as­set prices. Against this, the ne­olib­eral ar­gu­ment, so to speak, is two-fold.

First, there is a con­jec­ture har­bor­ing the pos­si­bil­ity that pri­vate eq­uity will di­ver­sify risk away. It ar­gues that pri­vate own­ers may dis­cretely trans­fer risk to coun­ter­par­ties – if mar­kets tend to be­come ‘com­plete’, that is if the ar­gu­ment for mar­ket-clear­ing at in­fin­ity can be cast away - mar­kets, who can man­age or oth­er­wise di­ver­sify away those risks they choose to forego, ar­guably be­com­ing a lower cost sub­sti­tute for tra­di­tional risk cap­i­tal. The ar­gu­ment aims at dis­cred­it­ing any trans­fer to the pub­lic in the pres­ence of con­strained in­ef­fi­ciency to rem­edy one spe­cific curse of in­com­plete as­set mar­kets. We know that with an in­com­plete span of as­sets, com­pet­i­tive mar­kets do not func­tion op­ti­mally even within the re­stricted sub­set of ex­ist­ing mar­kets. If this math­e­mat­i­cal ar­gu­ment can be so eas­ily over­turned, in­deed, if risk man­age­ment can sub­sti­tute for risk cap­i­tal with­out re­quir­ing a trans­fer of own­er­ship, then why go pub­lic at all? This first part goes against the whole lit­er­a­ture of gen­eral equi­lib­rium the­ory with in­com­plete mar­kets, re­stricted par­tic­i­pa­tion and all that of the 1985-2005 pe­riod.

Se­cond, there was Lehman, and the cri­sis, and the sec­u­lar stag­na­tion, and all that, re­mem­ber! We all un­der­stand that from a tech­ni­cal view­point many a fi­nan­cial sys­tem/ in­sti­tu­tion has en­tered the down­turn of the credit cy­cle with heav­ily in­vested as­set port­fo­lios, over­heated real es­tate mar­kets, pri­vate eq­uity/LBO and struc­tured credit prod­ucts en­tail­ing a num­ber of L-geared risks: longer du­ra­tion, lower credit qual­ity, larger po­si­tions, lever­ages, less liq­uid as­sets. It is also clearly un­der­stood that, de­spite lots of ex­am­ples cited above and that went at times right to the heart of ex­ist­ing lit­er­a­tures, peo­ple do not seem to have taken se­ri­ously myr­iad ways of ad­vis­ing against bear­ing po­ten­tially un­bear­able risks. Was it all about “fat tails”, i.e. non-lin­ear­ity and non-Gaus­sian­ity, etc., or was it also about the macroe­co­nomic frame­work, a frame­work that sheds its shadow to the fi­nance world via mon­e­tary and credit chan­nels? One, cycli­cal risks are low-fre­quency/high-im­pact events char­ac­ter­ized by their neg­a­tively skewed and “fat-tailed” loss dis­tri­bu­tions. This means in­vestors in­cur­ring such risk can ex­pect mainly small pos­i­tive events but are sub­ject to a few cases of ex­treme loss. Con­se­quently, they are dif­fi­cult to un­der­stand. The dif­fi­culty stems from two fac­tors. First, there is in­suf­fi­cient data to de­ter­mine mean­ing­ful prob­a­bil­ity dis­tri­bu­tions. Se­cond, and per­haps more im­por­tantly, in­fre­quency clouds haz­ard per­cep­tion. Risk es­ti­mates be­come an­chored on re­cent events. This leads to dis­as­ter my­opia, whereby we ig­nore

low-fre­quency, re­mote events. Overem­pha­sis on re­cent events can also pro­duce dis­as­ter mag­ni­fi­ca­tion im­me­di­ately fol­low­ing an ad­verse event. These facts lead to risk mis­pric­ing and the pro-cycli­cal na­ture of risk ap­petite. Two, the other prob­lem is cor­re­la­tion within a port­fo­lio; that is among the as­set re­turns that form it. In this re­spect, us­ing long-mem­ory data sets has been of no avail, since such a prac­tice only, per­haps, aug­ments the pa­ram­e­ter space since co-vari­a­tion ma­tri­ces can grow very large.

Then Li came up with a con­vinc­ing idea, sort of a Black-Sc­holes equiv­a­lent de­vice that all in­vest­ment houses could eas­ily use and hedge against port­fo­lio cor­re­la­tion risks. Li, a Chi­nese ex­change stu­dent orig­i­nally, who went on to pick up a PhD in ac­tu­ar­ial sci­ence from Canada, ended up in Wall Street. The Gaus­sian cop­ula he used al­legedly al­lowed fi­nan­cial in­sti­tu­tions to sep­a­rate mar­ginal dis­tri­bu­tions from their joint de­pen­dency. Not only cor­re­la­tion is con­toured within the as­set port­fo­lio, let us say a CDO, but also it is pos­si­ble to com­pute the de­fault time of a com­pany if the as­set value falls be­low some thresh­old value. Gone all the risks, gone the in­com­plete mar­ket hy­poth­e­sis which no longer mat­ters, and back is the cling­ing as­so­ci­a­tion that mar­kets can al­ways pro­vide the best so­lu­tion. John Geanako­p­los from Yale con­cisely put for­ward the the­o­ret­i­cal and, hope­fully, prac­ti­cal im­pli­ca­tion back in 1990: “the au­to­matic as­so­ci­a­tion of the ex­is­tence of com­pet­i­tive mar­kets with some kind of in­tu­itive no­tion of ef­fi­ciency in the mind­set of prac­tic­ing economists has now no the­o­ret­i­cal va­lid­ity what­so­ever.”

Prac­tice…

Now, both ar­gu­ments are in­cor­rect. Se­cu­ri­ti­za­tion serves noth­ing in terms of mit­i­gat­ing mar­ket and port­fo­lio cor­re­la­tion risks; a mere Gaus­sian cop­ula can­not han­dle all the com­plex­i­ties of as­set mar­kets; pri­vate eq­uity can­not pass for di­ver­si­fied pub­licly hold risk slices. Have we not yet learned – so painfully - that this kind of ir­ra­tional ex­u­ber­ance and mar­ket fun­da­men­tal­ism only leads to fi­nan­cial and real crises? Se­cu­ri­ti­za­tion is a bad idea in­deed, and it is ex­tremely bad es­pe­cially if you are pre-set to pack up all even­tu­al­i­ties in an­tic­i­pa­tion and mar­ket se­cu­ri­tized credit chunks to a cen­tral bank. This ar­gu­ment amounts to ac­cept­ing – so many years af­ter the cri­sis and ob­vi­ously with hind­sight - that as­set-backed se­cu­ri­ties that banks are chan­neled to is­sue may go sour with such a high prob­a­bil­ity that the CBRT has bet­ter buy them right away from day one. This is not only wrong on all counts, but it has a sig­nal­ing ef­fect that is self-de­feat­ing. Are we so des­per­ate to strongly in­cen­tivize banks and - wrongly - give them se­cu­rity guar­an­tees that go much be­yond what a well-de­signed credit guar­an­tee fund would en­tail, in­duc­ing them to ex­tend cred­its at all costs? Let us see.

Clearly growth is pred­i­cated upon do­mes­tic spend­ing of all sorts since ex­ports are of no help. This means, primo, pub­lic spend­ing, and we see the im­pact of that in the pri­mary bal­ance trends de­picted in the rel­e­vant graph, and sec­ondo trig­gered and pro­longed loan growth; we see that in the graph as well. Pub­lic banks lead and pri­vate banks fol­low, rem­i­nis­cent of the orig­i­nal in­tent of the Von Stack­el­berg-Zeuten game in the 1930s.

We have re­peat­edly claimed that im­pe­tuses and in­cen­tives of the sort ob­served in the last five years would suf­fice to carry GDP growth above three per­cent, and more likely above 3.5 per­cent eas­ily. That would be still sub­par growth. But now that the hit is al­most com­plete, only near zero growth can be pro­jected. We also claimed, back in 2017 for in­stance, that based on base ef­fects, that the ini­tial mo­men­tum pro­vided by ac­cel­er­ated pub­lic spend­ing-cum-bank lend­ing would suf­fice for achiev­ing that end. How­ever, these suc­cess­fully used – and warn out - mech­a­nisms are be­ing re­sorted to as if in a re­peated game that will never end. In as much as we find the ex­is­tence of risk-mit­i­gat­ing guar­an­tee funds for SMEs, small ex­porters, and agri­cul­tural firms sound, the­o­ret­i­cally and prac­ti­cally, help­ful, we see se­cu­ri­tized as­set tranches bound to be plugged in the bal­ance sheet of the cen­tral banks even be­fore their is­suance, a telling rep­e­ti­tion of the post-2008 sce­nario glob­ally speak­ing at a time when the world is re­volv­ing around a new at­trac­tor. The cri­sis in its first and nar­row phase was over al­ready in 2013 when the Fed sig­naled the end of as­set pur­chases, and now the sec­u­lar stag­na­tion is over, or about to be. Is the sit­u­a­tion so des­per­ate the pub­lic is pre­ma­turely called in the game in a way that would as­tound all bank­ing an­a­lysts and mon­e­tary economists?

Rough ride ahead…

There also is the pos­si­bil­ity that given all sorts of moral haz­ard is­sues, a cen­tral bank-spon­sored pri­vate bank credit se­cu­ri­ti­za­tion – ob­vi­ously a CDO since even­tu­ally the process is repet­i­tive and self-gen­er­at­ing - will re­sult; liq­uid­ity will also be an is­sue. If you pro­vide TRY liq­uid­ity in ex­cess of spot de­mand, there is a very good chance that this money will end up in FX de­mand, in ei­ther the short or long end of the curve. Liq­uid­ity is a short-term is­sue by na­ture, and many short-lived mon­e­tary phenom­ena will en­sue from its ex­cess. Con­sider the cur­rent 49 per­cent FX de­posit-to-M2 ra­tio of lo­cal res­i­dents. It has been trend­ing up since the ‘stress­ful sum­mer’ of 2013. Re­cently, af­ter a nor­mal­iza­tion and back re­ver­sion episode, that ra­tio is head­ing south anew. Now, 50 per­cent is the key thresh­old for this key in­di­ca­tor, above which dol­lar­iza­tion could not only ac­cel­er­ate but be­come per­sis­tent. We are just hov­er­ing around ‘the full dol­lar­iza­tion ahead’ sign­post.

Se­cu­ri­ti­za­tion through the Cen­tral Bank isn’t a good idea. It may be a dan­ger­ous idea that has ram­i­fi­ca­tions and reper­cus­sions that could go much deeper than we may be ready to ad­mit.

Newspapers in English

Newspapers from Turkey

© PressReader. All rights reserved.