FinTech Summary
This summary will cover all the relevant information from the course FinTech and the
transformation of financial services. It will first cover the content of the 8 lectures provided
by mr. Schmitz, and then all the six guest lectures. At the end the group presentations are
also discussed. Keep in mind that the summaries of these topics do not cover the
presentations fully. Since a lot of ‘trivial’ information is not going to be relevant information
for the exam, I kept it out for a large part. Only the in depth conclusions are still covered.
Written by Billy den Otter
,Lecture 1 Introduction to FinTech
What is FinTech?
- Academic definition: Fintech is a new financial industry that applies technology to
improve financial activities.
- Business definition: fintech companies are businesses that leverage new technology to
create new and better financial services for both consumers and businesses. It includes
companies of all kinds that may operate in personal financial management, insurance,
payment, asset management etc.
Disruption of existing industries: the first wave
- new technology and the introduction of the internet made some businesses obsolete
- especially businesses that deliver information on physical media and forward information
- newspaper industries
- DVD and music stores
- many of these businesses didn’t even think they were in the information sector, like the
music industry not realising they were also selling information on physical media.
- smartphones picked up popularity and started replacing physical features with apps.
Disruption of existing industries: the second wave
- growing internet usage, more scope and economies of scale
- growing information advantage of tech firms over other actors in the economy.
- Business models at risk:
- rely on customer trust or are based on knowing customer characteristics
- provide services which can be substituted by an informed digital agent
- this made another group of businesses obsolete
- travel agencies and hotel businesses
- cab drivers, especially call centres for them
- business models currently at risk of disruption are information-centric or trust-based
models of intermediation
However, this ‘disruption’ is not that surprising as people might think, but rather the product
of predictable and long development. It is mostly a problem of businesses not realising they
are at risk.
Moore’s law first long-term trend. Moore’s law shows the number of transistors on
integrated circuit chips. See the slide what it compares to what.
- there is a free miniaturization dividend for each round of miniaturization. The number if IC
transistors doubles every 18 months.
,Lecture 2 Diamond Dybvig model and asymmetric
information
Banks are informed intermediaries. They bridge information asymmetries between agents.
E.g: banks have information about riskiness of borrowers and lend money to them, basing
their interest rate on the riskiness.
The key thing banks are about is described in the Diamond and Dybvig model.
- imagine a solitary person living on a deserted island
- we assume:
- three time periods, 0 1 and 2.
- consumers have an endowment of one unit of good in time zero, and no
endowment in future periods. Think of a bunch of mangos
- Certainty about the world but unnecessary about own preferences
consumption at t = 1 or t = 2.
- He questions if he should eat the mangos (consume at t=1) or if he should bury the mangos
to start a tree and eat more mangos in the future (consume at t=2)
- if you consume at t=1 you cannot consume at t=2. This works both ways: you cannot
consume if you decide to invest in growing a tree.
- there will be less return if you consume right now.
There are two strategies:
- storage: keep access to the endowment
- keep the mangos for consumption between t=1 and t=2
- convenient, but not productive. It yields in nothing extra
- like putting spare money in a sock
- investment: give up access to endowment in t1 to get higher return in t=2.
- illiquid asset which matures in t=2
- Every unity of the long asset gives a payoff of R>1 at t=2
- every unit of the long asset gives L<1 if liquidated early at t=1
- think of planting the mangos gives a tree that yields more mangos (R>1), but digging
them up early makes them taste worse than if you just kept them in storage (L<1)
At t=0, he doesn’t know whether he wants consumption at t=1 or t=2
- with probability π1 they are impatient, and want to consume early at t=1
- with probability π2, they are patient and consume in t=1
Assuming concave utility u(.) with u’(.) > 0 and u’’(.) < 0
expected utility is the probability you do an action times the return of that action.
expected utility from consuming c1 in t=1 and c2 in t=2 will give you
EU = π1u(c1) + π2u(c2)
, Say the consumer on the island invests I in the long asset and 1-I in the short asset. If the
consumer ends up being impatient, he will consume c1 = LI + (1 – I) in t=1, liquidating the
long asset plus consuming all short asset payoffs. If the consumer ends up being patient, he
will consume c2 = RI + (1-I) in t=2, liquidating the long asset in time two plus all the short
assets payoffs.
You don’t know if you will choose t=1 or t=2 consumption, so you will have to mix.
Because the maximum investment is 1 and because L<1, any choice of I implies that c 1 ≤ 1
and c2 ≤ R with one of these inequalities holding strictly, you can see that consuming c1 = 1
and c2 = R is not possible.
mixing will always make you worse off than committing to one goal.
Now imagine there are also other people on the island expanding the model
- impatient consumers can sell a bond in t=1 for a price p to patient consumers instead of
liquidating the illiquid asset at a loss.
they can trade with each other.
- they will consume early what they get from the bond, and repay the bond with the payoff
of the long asset at t=2
- patient consumers will use all their short assets to purchase bonds in t=1 for the price of p
mangos from impatient consumers (1-I)/p. The logic behind is, is that all they don’t invest p. The logic behind is, is that all they don’t invest
and can be shorted to impatient people because they don’t consume it themselves.
- impatient consumers consume c1 = pRI + (1 – I)
- patient consumers consume c2 = RI + (1 – I)/p. The logic behind is, is that all they don’t invest p
- because of this, c1 = pc2
- the utility of consumers increases in I if pR > 1 and decreases in I if pR < 1.
optimal where pR = 1 and thus where p = 1/p. The logic behind is, is that all they don’t invest R
- if we insert p = 1/p. The logic behind is, is that all they don’t invest R in the model:
- impatient consumers consume c1 = 1/p. The logic behind is, is that all they don’t invest R *RI + (1 – I) , or c1 = 1
- patient consumers consume c2 = RI + (1 – I)/p. The logic behind is, is that all they don’t invest 1/p. The logic behind is, is that all they don’t invest R , or c2 = R
- this proves that this system is strictly better than autarky, since c1 = 1 and c2 = R was not
reachable in autarky but can be reached in a trade situation.
Next, we add social planners to the model. There is a social planner on the island who
allocates resources to maximize E(U) of all consumers on the island.
So max. π1u(c1) + π1u(c2)
maximizing gives π1c1 = 1-I and maximizing π2c2 = RI, or c1 = (1-I)/p. The logic behind is, is that all they don’t invest π1 and c2 = RI/p. The logic behind is, is that all they don’t invest π2
substituting this into the max function gives
max. π1u((1-I)/p. The logic behind is, is that all they don’t invest π1) + π1u(RI/p. The logic behind is, is that all they don’t invest π2)
First order condition: taking the first derivative and setting it to zero in order to find the
maximal optimum of a function.
u’(c1*) – Ru’(c2*)
