Structured Finance Flashcards
What is structured finance ?
Mainly about pooling of economic assets and issuing prioritized claims against collateral pools like :
- Senior tranche
- Mezzanine tranche
- Junior tranche
What can structured finance achieve ?
- Diversification
- Broaden pool of potential investors
- Improve liquidity
- Lower transaction costs
- Extend credit to many more customers than when banks would have to hold all loans on BS
- Lower borrowing costs for corporations
What are many institutions demanded and who are they ?
Explicitly required to hold certain share of AAA-rated securities
- Insurance companies
- Pension funds
- Mutual funds
How did Basel II and other regulations render AAA-rated securities and why ?
Highly attractive from capital requirement perspective, because :
- Zero risk weight
- AAA → pledged fully as collateral in day-to-day repo lending
→ however, natural AAA very limited
What are the two types of investors ?
- Pension fund : lot of capital but risk averse + willing to hold investment-grade securities
- Risk neutral investor : capital constrained + only invest one unit
What happens when structured finance products are absent ?
Production inefficiently low
What happens if bank buys two bonds to create collateralized debt obligation ? (CDO)
Characteristics :
• One senior and one junior tranche
- Defaults after one year with p = 5%
- Bonds = independent with each other
- Rating of CCC+
- Junior tranche defaults if at least one bond defaults with probability : 1 – (1-p)^2
- Senior tranche defaults only if both bonds default with probability p^2
→Senior tranche = much safer than underlying asset + achieves investment-grade rating A-
→ sell senior tranche to pension fund and junior to risk-neutral investor
What is the economic benefit of structured finance ?
Welfare-enhancing as allows to create products matching demand side’s risk preferences and capital constraints
→ highly desirable
What are the two main pitfalls of structured finance ?
• Rating error : risk of understimating loss distribution of underlying asset pool
o Too many securities AAA-rated
o Compounded by more complicated products
• Pricing error : conditional on ratings alone not sufficient
o Senior tranches tend to default in economy’s worst states
o Important systemic component should be priced into asset but it’s not
What do the two main pitfalls create ?
Enormous profits for issuers of structured products
How do we calculate the probability PD (x,p= that tranche x defaults ?
There are 100 bonds
PD(x,p) = 1-∑(100;i) p^i (1-p)^(100-i)
where 100;i is the binomial coefficient
How is it possible to create a CDO-squared ?
By computing a CDO of another CDO
What is the problem with CDOs and especially CDO-squared ?
- Extremely sensitive to changes in default probabilities of underlying assets
- Default function highly non-linear
→ Rating errors hug effects, especially for CDO-squared
What are the main factors that contributed to rating error regarding mortgage-based structured products ?
• Extremely high demands for accuracy to rate structured product as AAA
o Uncertainty regarding some key estimates such as default rates of different mortgage products neglected
• Wrong key parameters forecasts
o Assumption based on post-war experience (house prices never decline)
• Rating structured products requires to take correlation between huge set of assets into account
o Rating agencies used to rate bonds of individual issuers lacked necessary experience
What did Coval et al. agrue ?
There is a huge systematic risk in senior tranches of structured products
What are the important difference between risk of individual bond and structured product made up of bonds ?
Economic conditions → small role for risk assessment of individual bond whereas it plays key role for prediction mass defaults of asset class.
Why are senior tranche so problematic in structured producs ?
Default in worst states of economy, marginal utility consumption highest:
- Senior tranches should command risk premium over equally rated individual bonds
- Using same rating-scale for structured products and individual bonds is misleading
What was wrong with structured finance ?
Ability to create nearly safe products = overstated
- Rating error → underestimated probabilities of default
- Pricing error concealed senior tranches default in worst states of economy
→ Correcting for rating and pricing error will greatly reduce number of AAA-rated securities created with given pool of assets
What is the puzzle that extraordinary boom and bust of MBS finance posed ?
Why mkt participants :
- ignore rating + pricing issues associated with MBS products
- react very strongly to first signs of declining house prices
What did Gennailoli et Al present ?
Behavioral explanation based on “local thinging” = agents’ inability to take all but most likely states into account.
What are Gennaioli et al. main intuition ?
- Simple setup with two types agents
• Investors = extremely risk avers + only interested in safe (AAA) securities which are scarce
• Financial intermediary = risk neutral + can create substitutes for safe securities by trancheing risky assets - Investors + financial intermediary both local thinkers
• Only focus on booms and downturns but ignore risk of severe recession
→ Financial intermediary issues too many structured products safe only in booms & downturns but not recession
→ Bad news makes agents suddenly aware of risk of recession
→ Flight into safe assets + sell off of structured products
What is Gennaioli et al.’s model structure ?
• Three periods
• 2 assets :
o A = risky pays y(i) with probability π(i) with i = g (growth), d (downturn), r (recession) s.t
Y(g) > 1 > y(d) > y(r)
π(g) > π(d) >= π(r)
o B = safe and pays R > 1 for sure
• Assets = only way to transfer wealth to period 2
Who are Gennaioli’s two agents ?
• Investors : initial wealth w and infinite risk aversion
o U = C + C(1) + θmin(C(2g), C(2d), C(2r))
o θ > 1 (investors like to consume in period 2)
• Financial intermediary : owns both assets and maximizes expected profit over all periods
What are the gains from trade in Gennaioli’s model ?
- Investors prefer to consume in period 2 (θ > 1), FI sells claims on B and A to allow transferring wealth
- Investors prefer safe claim on B due to risk aversion
What is Gennaioli’s model timing ?
1 : t = 0, trading of claims at prices p(b) and p(a) takes place
- t = 1, all agents observe signal s with two possible realizations
• If s = û : probability of growth higher
• If s = u: probability growth lower & recession more likely than downturn
→ after observing signal, new trading at prices p(b1) and p(a1) - t = 2, assets pay off
What are the three assumptions of competitive markets ?
- Rational expectations with traditional claims
- Rational expectations with financial innovation
- Local thinking and financial innovation
What is the model under rational expectations with traditional claims ?
• Investors’ reservation price for buying
o Safe bond = θR
o Risky assets = θy(r)
• Financial intermediary’s reservation price for selling
o Safe bonds = R
o Risky assets = E(y)
→ Assumption that investors value risky assets always less than intermediary ( θy(r) < E(y|u) ) = only trading in safe bonds
→ financial intermediary keeps all risky assets
When does the model under rational expectations with traditional claims achieve equilibrium in t = 0 ?
Investors wealthy and “shortage” of safe bonds
What happens after the signal is observed in t = 1 in the model under rational expectations with traditional claims ?
Nothing happens to portfolios and consumption levels :
- Price of safe bonds stays at p(b) = θR
- Only price of risky asset p(a) fluctuates as expectations adjusted to realized signal s
What is the model under rational expectations with financial innovation ?
• Shortage of safe bonds → social benefits from creating additional safe assets
o Investors value consumption in t = 2 more than intermediary (θ > 1)
o Investors infinitely risk averse → consumption in t = 2 certain
• Financial innovation → form of structured product intermediary creates by tranching risky assets’yield y
o Intermediary sells senior tranches that guarantee payoff R in t = 2 and keeps remaining junior tranches
o Maximal amount of senior tranches : f = y(r)/R
o Senior tranches = perfect substitutes for safe bonds B
When does the model under rational expectations with financial innovation achieve equilibrium in t = 0 ?
investors wealthy + absorb all safe assets at price θR
What happens after the signal is observed in t = 1 in the model under rational expectations with financial innovation ?
Nothing happens :
- Price of safe assets (bonds + structured product’s senior tranches) stays at p(b) = θR
- Price of risky asset p(a) fluctuates as expectations adjusted to realized signal s
→ Reason : Asset A structured very conservatively so that senior tranches pay R for sure even in case of recession
What is the first result of Gennaioli’s model ?
Structure finance can ve welfare-enhancing in world with rational expectations
What is the model under local thinking and financial innovation ?
All agents = local thinkers and take only two most likely states into account :
• T = 0 : most likely states = g and d
• T = 1 : states’ likelihood depends on signal s
o If s = û : g and d remain most likely states
o If s = u : g and r = two most likely states
• Key idea : agents neglect possibility of severe recession r until observe bad signal u
What is the maximum amount of senior tranches intermediary can issue ?
f(L) = y(d)/R > y(r)/R = f
→ agents ignore possibility of server recession, believe risky asset A yields at least y(d)
→ intermediary can now issue too many senior tranches that perceived as perfect substitutes for safe bonds B
→ local thinking causes rating error
When does the model under local thinking financial innovation achieve equilibrium in t = 0 ?
investors not wealthy enough to absorb all safe assets at price θR
What happens after the signal is observed in t = 1 in the model under local thinking and financial innovation ?
Signal s’s realization determines what happens
• If s = û
o Structured product’s senior tranches perceived as perfect substitutes for safe bonds B
• If s = u : things go terribly wrong
o Agents suddenly realize recession r possible and senior tranches may only pay : (y(r)/y(d))R < R
o No longer perceive senior tranches as safe assets
o Mkts for senior tranches and safe bonds B separate
What happens once mkts for senior tranches and safe bonds B separate under local thinking and financial innovation ?
• Investors’ reservation price for trading
o Safe bonds is θR
o Senior tranches is θ(y(r)/y(d))R
• Financial intermediary’s reservation price for trading
o Safe bonds = R
o Senior tranches is [P(L)(y(r)|u)(y(r)/y(d)) + P(L)(y(g)|u)]R = w(L)R
What if θ(y(r)/y(d))R > w(L)R ?
• Investors value senior tranches higher than intermediary
• Not trading takes place, but prices must adjust to
o P(b1) = θR > p(B)
o P(ST1) = θ(y(r)/y(d))R < p(b)
o Investors = indifferent between buying or selling
• Flight into secure assets: price of safe bonds p(b1) raises and price of senior tranches p(ST1) plunges
What if θ(y(r)/y(d))R < w(L)R ?
- Investors value senior tranches less than intermediary
- Intermediary eager to buy back senior tranches
- Intermediary may not be wealthy enough to buy back all previously issued senior tranches
- Any case, price of safe bonds p(b1) raises and price of senior tranches p(ST1) plunges
What is the second result of Gennaioli’s model ?
Mkts for presumably safe structure products = fragile
- When investors get surprising news about unattended risks, sell off false substitutes and flle into traditional safe assets
- Sell-off of structured products may come in addition to fire sales driven by high leverage
- Overissuance of structured products = source of risk
What is the third result of Gennaioli’s model ?
When risk averse investors sell off structured products, risk neutral intermediaries may be willing to buy them back
- Explain why banks, after having been recapitalized wanted to buy back MBS instead of providing additional credit
- Intermediaries wealth provides insurance vs collapsing prices of structured products
