►
From YouTube: Pillar Series: An intro to DeFi
Description
Learn why DeFi matters and how it's changing the financial landscape. Rowland dives into AMMs, lending, borrowing, decentralized exchanges, and more. This introduction should give you the tools to dig into decentralized finance. You will have a better understanding of the real problems DeFi is solving and maybe walk away with inspiration to build something yourself!
A
Hello,
everybody
welcome
to
gorick's
pillar
series.
A
So
as
a
reminder,
I
I'm
roland
grouse
product
leader
at
agorik
and
we
do
a
pillar
talk
once
roughly
every
month
or
so
usually
around
topics
that
are
a
little
outside
of
our
normal
speaking
speaking
engagements,
a
little
less
technical,
maybe
more
market
focused
and
today
I'm
going
to
do
an
intro
to
defy
so
you
know
our
hoped
for
audience,
for
this
talk
are
developers
or
people
who
have
observed
the
d5
ecosystem
from
the
outside,
but
either
have
only
started
to
wade
in
or
are
potentially
learning
about,
some
of
the
core
concepts
still
so,
if
you're,
a
crypto
native,
if
you're
aping
into
unaudited
contracts
that
are
30
minutes
old,
this
talk
is
probably
not
for
you.
A
We'll
have
content
for
for
you
later
on,
but
I'd
love
to
take
you
through
through
this
intro,
and
you
know,
during
the
talk,
please
send
questions
in
you
know
I
see
a
number
of
you
in
youtube,
so
I'm
sort
of
managing
everything,
I'm
I'm
doing
the
the
youtube
stuff,
I'm
I'm
doing
the
deck
so
I'll
be
back
and
forth,
but
would
love
to
take
live
questions?
A
If
you
have
them
so
with
that,
let's
talk
about
our
goals,
so
I
I
said
already:
this
is
an
intro
to
d5
talk
and
what
I'd
like
to
do
to
start
with
is
frame
the
d5
movement
talk
about
what
are
the
problems
it
really
solved?
Is
it
solving
real
problems,
or
is
it
just
a
bubble?
A
Talk
about
some
of
the
use
cases
that
are
involved
in
defy
and
then
we're
going
to
dig
into
two
specific
use
cases,
lending
and
automated
market
makers,
and
my
goal
with
that
is
really
just
to
scratch.
The
surface
of
some
of
the
core
concepts
penetrate
some
of
the
language
defy,
as
you
know,
has
some
of
its
own
intricate
language
around
a
certain
mechanics.
A
So
I'd
like
to
talk
talk
through
some
of
those
and
then
suggest
a
few
future
directions
for
those
for
those
areas
and
then
beyond
that
talk
about
what
it
might
look
like
to
launch
a
d5
product,
some
things
to
think
about,
if
you're
looking
to
do
that
and
then
how
agora
can
help
and
why
we
think
agoric
is
a
good
place
to
be
doing
that
if
you
are
launching
a
project
all
right.
A
So
with
that
again
happy
to
take
questions
as
we
go,
I
will
stop
at
the
end,
but
please
please
stop
me
and
I'll
I'll
go
back
or
elaborate
on
additional
concepts.
A
All
right!
So
what's
happened
with
d5
over
the
past
several
years
it
this.
This
graph
shows
the
ethereum
locked
in
defy
contracts,
which
is
really
just
a
measure
of
how
much
attention
is
in
this
area,
how
much
how
much
funding
has
really
gone
in.
So
this
is
not
east
that
has
been
sold
for
d5.
This
is
if
it
is
being
used
to
lend
out,
or
do
some
of
the
other
things
that
we're
going
to
talk
about
in
the
use
cases
later.
A
You
see
this
chart
often
shown
in
dollars,
and
I
like
to
show
it
as
eve
just
to
sort
of
pull
out
the
the
dollar
value
going
up
as
part
of
it,
but
even
really
the
reality
is
somewhere
in
between
that
right.
So
as
of
this
point,
43.5
billion
dollars
worth
of
ethereum
is
locked
in
various
d5
contracts
and
that
has
grown
considerably
over
the
last
year
and
really
what
defy
did
was
it
made
idle
capital
productive?
So
you
know
people
have
been
holding
cryptocurrency
since
2010
and
in
2017
2018
2019.
A
A
lot
of
capital
really
came
into
the
system
and
I'm
I'm
showing
here
a
specific
section
of
the
total
crypto
market
cap
in
the
2018
to
mid
20,
20
20
20
time
frame,
and
you
can
see
that
somewhere
between
100
to
350
billion
dollars
worth
of
crypto
assets,
we're
sitting
largely
idle
and
what
we'll
show
and
what?
What
really
happened
was
defy
allowed
users
holding
their
crypto
assets
to
do
something
productive
with
them
to
earn
a
return
on
them,
and
that
is
really
sort
of
what
what
kicked
off.
A
You
know
what
what's
being
called
d5
summer
last
year
and
and
the
continuing
rise,
so
to
give
my
definition
for
what
what
a
d5
product
really
is
it,
and
this
is
just
personal,
so
feel
free
to
disagree.
But
for
me,
a
d5
product
is
something
that
is
a
permanently
deployed,
credibly,
neutral
application,
allowing
users
to
permissionlessly
exchange
assets
or
risk.
So
you
know
I've.
I've
often
heard
finance
describe
as
all
finances,
effectively
the
exchange
of
assets
or
the
exchange
of
risk
and
for
d5.
A
As
long
as
the
underlying
network
that
they
are
running
on
is
still
live,
those
contracts
still
exist,
the
functions
can
be
called,
the
logic
will
execute,
so
these
are
permanently
deployed
and
the
idea
would
be
incredibly
neutral
and
in
the
reality
of
that
is
complex,
but
over
time
many
of
these
applications
make
it
get
to
the
point
where
there's
no
governance
at
all
and
as
a
user,
you
know
that
there's
nobody
there's
no
third
party
that
can
interfere
with
the
the
running
of
a
specific
contract.
A
Okay,
so
let's,
let's
get
to
one
one
reason
defy,
is
different
and
why
it
has
captured
so
much
attention
and
and
that's
really
that
the
assets
themselves
or
the
projects
themselves
generate
cash
flow
and
that's
typically,
it's
done
in
several
ways
and
the
recipients
of
those
cash
flow
might
be
different
people,
so
they
may
be
holders
of
a
token.
They
may
be
actual
participants
or
users
of
the
of
the
product.
A
That
aren't
holding
a
specific
token,
but
fees
that
are
generated
based
on
the
use
of
these
protocols
do
flow
to
users,
and
that
is
something
that
you
know
when
you
go
in
and
talk
to
actual
capital
markets.
People
they've
had
a
lot
they've
had
a
hard
time
understanding
how
to
value
crypto
projects,
because
it's
so
different
from
what
they're
used
to.
A
Typically,
if
you
look
at
a
typical
l1,
crypto
asset
valuing
that
has
been
difficult,
but
you
can
go
and
say,
look
at
uniswap
and
look
at
the
cash
flows
that
are
being
generated
over
time
and
they
can
plug
that
into
a
model
and
generate
a
value
for
it
right.
So
cash
flows,
equal
valuation
and
d5
has
started
driving
that
driving
that
reality
and
and
made
institutional
investors
wake
up
as
a
result.
A
Okay,
so
we
don't
have
time
to
talk
about
all
these
d5
use
cases.
This
would
this
would
be
a
300
slide
deck,
but
I
do
want
to
go
into
a
little
bit
more
detail
about
lending
and
dexes
or
what
you
know.
I'm
specifically
talking
about
automated
market
makers
in
that
case,
but
it's
important
to
know
that
defy
actually
does
approach
a
whole
bunch
of
different
use
cases
so
stable
coins.
A
Many
of
you
are
already
aware
of
those
and
and
bill
did
a
pillar
series
talk
on
stable
coins
a
couple
months
ago,
synthetic
assets,
derivatives
and
then
an
emerging
class
of
asset
managers.
So
you
know,
put
your
put
your
assets
in
one
place
and
then
they
will
do
all
the
complicated
logic
for
you
to
figure
out
where
you
can
get
the
best
return.
That
sort
of
thing
there's
more
categories
than
I'm
listing
here,
but
this
is
a
useful
way
to
group
different
d5
d5
projects.
A
All
right,
so,
let's
dig
into
lending
a
little
bit
and,
as
I
said
at
the
beginning,
len
the
goal
here
really
is
not
to
make
sure
that
you
understand
everything
that
you
possibly
could
about
how
lending
projects
are
built.
A
But
I
want
to
go
deep
enough
so
that
you
understand
some
of
the
mechanics
talk
a
little
bit
about
terms
that
you've
heard
in
the
industry
so
that
they
connect
to
something
when
you
hear
them
again
and
then
give
you
a
couple,
you
know
terms
to
google
or
things
where
you
could
explore
further,
if
you're,
if
you're
interested
so
to
start
with
what
are
the
actual
use
cases
for
borrowing
right,
you
know
what
problems
is
this
actually
solving
for
users,
and
I
I
sort
of
summarized
this
into
four
the
first
in
and
this
I'm
I'm
guessing.
A
This
was
the
primary
one.
To
start
with,
is
it
allows
users
to
get
liquidity
on
their
crypto
assets
in
dollars,
for
example,
without
having
to
sell
them
right?
So
when
you
sell,
you
know
if
you're,
holding
ethereum
or
any
other
crypto
asset,
and
you
sell
it.
Not
only
do
you
lose
exposure
to
that
asset,
you
know
if
it
keeps
going
up,
you're
no
longer
part
of
that,
but
you
also
potentially
trigger
capital
gains
right
and
well,
you
definitely
trigger
capital
gains.
A
You
may
trigger
short
term
depending
on
how
long
you've
held
it,
and
that's
it
an
important
thing
that
users
want
to
avoid
so
by
borrowing
against
their
crypto
assets.
They
can.
Actually.
You
know
I
could
borrow
usdc
in
the
flow
that
I
show
here.
Bar
usdc,
send
it
to
coinbase,
convert
usdc
to
us
dollars,
send
to
my
bank
account
and
then
live
my
life
right,
so
I
can
actually
get
get
dollars
out
of
the
growth
of
my
crypto
assets
so
that
that's
use
case.
A
One
use
case
two
which
again,
I
think
is
driving
a
lot
of
borrow
demand
is,
is
levering
up,
and
this
is
a
standard
thing
that
in
in
finance
you
are,
you
often
see
and
apologies
if
my
my
key
based
notifications
are
coming
through,
but
leveraging
up
means
taking
a
a
borrowed
position
and
increasing
and
increasing
your
holdings
right.
So,
in
this
case,
I
would
supply
ethereum
to
a
lending
protocol,
borrow
usdc
against
my
ethereum
and
then
use
that
usdc
to
buy
more
ethereum,
so
I'm
taking
on
more
risk.
A
I
now
have
a
borrowed
position
and
I'm
a
little
bit
more
exposed,
but
this
is
a
way
for
for
users
to
get
more
exposure
if
they're
extremely
bullish.
On
a
particular
token,
the
the
third
use
case
is
native
shorting
of
crypto
assets,
so
in
in
standard
finance,
you
know
if
you're,
holding
a
stock
and
say
you're
doing
it
through
fidelity
or
something
like
that.
A
Fidelity
will
lend
that
stock
out
to
anyone
that
wants
it
for
a
certain
rate
and
allow
them
to
sell
it
and
buy
it
back
lower
or
that's
their
hope
anyway,
to
profit
from
the
stock
going
down
and
that's
those
are
the
internal
mechanics
of
how
shorting
works,
and
these
lending
and
borrowing
protocols
allow
users
in
the
cryptosphere
to
natively
short
something.
So
if,
for
example,
I
think
ethereum
is
going
down,
I
could
borrow
ethereum
from
a
lending
protocol,
sell
it
in
the
open
market
for
usdc,
hold
that
usdc
for
some
period
of
time.
A
And
then,
if
I'm
right,
when
I
buy
eve
back
at
a
lower
price,
I
have
to
spend
less
usdc
for
it.
I
can
return
the
eth
repay
my
debt
and
I
keep
the
difference
right
so
that
that
is
the
mechanics
of
how
shorting
work,
shorting
works
and
then
the
fourth
I've
grouped
into
areas
that
I'm
not
going
into,
which
are
crypto
specific
use
cases
so
flash
loans,
that's
something
I'll
talk
about
briefly.
A
In
other
places,
governance
which
really
is
sort
of
more
governance,
attacks,
there's
additional
things
that
that
you
might
want
to
use
this
for
okay.
So
let's
talk
about
the
mechanics
of
how
how
these
lending
protocols
actually
work-
and
the
the
first
thing
to
think
about
is
that
lending
crypto
for
in
d5,
for
a
number
of
reasons
is
a
peer
to
pool
model
and
what
I
mean
by
that
is
as
a
lender
as
somebody
that
is
supplying
assets
to
the
protocol.
You're,
not
choosing
specific
borrowers.
A
You
are
lending
to
a
pool
of
capital,
so
if
I'm
lending,
if
I,
if
I
want
to
lend
my
ethereum
out,
I
am
lending
ethereum
to
a
pool
of
ethereum-
that's
grouped
with
all
the
other
lenders
right
and
similar
when
I'm
borrowing
ethereum,
I'm
not
borrowing
from
somebody
specific
I'm
borrowing
from
that
pool
and
that's
an
important
dynamic
that
that's
true
throughout
d5
for
a
few
reasons,
but
it
does,
it
does
factor
in
here.
A
The
the
second
thing
to
think
about
is
that
all
of
these
so
far,
all
of
the
primary
lending
and
borrowing
protocols
are
over
collateralized,
which
means
that
if
I
want
to
borrow
something
from
the
protocol,
I
need
to
have
at
least
as
much
of
that
I
need
to
have
at
least
as
many
assets
supplied
to
that
protocol,
and
so
that
means
that,
for
example,
you
know
in
your
normal
life,
you
may
want
to
borrow
to
go
to
school
or
something
like
that.
A
That
is
not
an
over
collateralized
loan
right,
you
are,
you
are
borrowing
against
your
future
earnings
and
what
you're
actually
collateralizing
effectively
is
trust
that
you
are
going
to
be
able
to
repay
that.
In
the
future,
so
what
that
and
we'll
talk
about
the
mechanics
of
liquidations
and
things
like
that
that
relate
to
the
over
collateralization
and
in
a
later
slide,
and
then
there
are
some
supporting
factors
that
that
help
these
protocols
operate
so
price
oracles
are
important.
A
There
is
a
separate
group
of
liquidators
which
again
we'll
talk
about
and
then
beyond
that
and
I've
taken
this.
This
screenshot
from
ave
they're
integrated
applications,
so
external
protocols
might
be
choosing
to
borrow
and
lend
assets.
A
Okay
and
and
just
to
show
you
what
a
ui
might
look
like
for
one
of
these
applications.
So
I
I,
though,
the
last
slide
was
from
ave.
This
one
is
from
compound
two
of
the
leading
lending
protocols.
A
You
can
see
here
that
I've
got
two
two
markets,
so
they
show
supply
markets
over
here
and
borrow
markets
over
here,
and
there
are
different
rates
right.
So,
if
I'm
supplying,
in
this
case
basic
detention
token,
I'm
only
earning
a
1.3
percent
return
on
that
per
year.
If
I'm
borrowing
it,
I'm
borrowing
it
at
a
much
higher
rate,
8.52
and
we'll
we'll
talk
about
why
those
rates
are
are
so
different
in
the
meantime.
A
But
part
of
that
is
there's
a
spread
that
the
protocol
itself
earns
and
that
gets
back
to
these
protocols
earned
fees.
It's
real
cash
flow
that
is
understandable
to
institutional
investors
as
a
user.
I
may
have
supplied
multiple
assets
down
here
and
that
will
all
go
into
my
supply
balance
and
based
on
the
over
collateralization
requirements.
That's
going
to
turn
into
a
borrow
limit
for
me,
so
I
can
borrow
a
certain
amount
and
below
that
amount.
A
I'm
fine
and
it'll
show
me
it'll,
show
me
how
how
close
I
am
or
how
at
risk
I
am
to
being
liquidated,
but
that's
sort
of
how
it's
structured
and
again,
you
know
I
don't
want
to
go
too
deep
into
this,
but
just
to
give
you
a
sense
of
what
this
looks
like
for
one
particular
application
and
again
this
is
just
a
given
front
end
for
it.
There
could
be
multiple
front
ends.
A
Okay,
so
I've
mentioned
liquidations
a
few
times.
So
let's
talk
about
how
that
actually
works
and
again
this
is
from
the
obve
documentation
in
this
case.
There's
a
loan
here
that
I'm
showing
on
the
left
side
and
the
collateral
that's
required
for
that
loan.
Again,
it's
over
collateralized,
so
there's
more
required
in
collateral
than
there
is
for
the
loan,
and
this
is
denominated,
for
example,
in
u.s
dollars.
So
you
know
I
might
be,
the
collateral
might
be
ethereum.
A
The
loan
might
be
usdc,
it
can
be
different
assets,
but
you
convert
both
to
one
specific
asset
in
this
case
usd
to
understand
the
difference
in
relative
value.
So
this
line
here
is
they're,
calling
it
the
liquidation
value
here.
Sometimes
that's
called
liquidation
threshold
or
minimum
collateralization
ratio.
You
see
these
terms
a
lot.
They
all
effectively
mean
the
same
thing,
which
is
that
when,
if
the
loan,
if
the
collateral
value
falls
below
this
line,
then
a
liquidation
event
occurs
and
a
liquidation
event.
A
Is
the
the
protocol
controls
the
collateral
that
you
have
in
the
system
and
is
able
to
sell
it
on
your
behalf?
So
it
goes
and
sells
some
portion
of
your
collateral
to
pay
back
your
loan
and
that's
how
that's
how
the
protocol
is
able
to
make
sure
that
it
doesn't
get
taken
advantage
of
right.
You
know
the
whole
goal
of
these
protocols
is
to
be
permissionless
to
allow
anyone
to
take
a
loan
out,
and
they
don't
even
know
who
it
is
the
way
they
manage.
A
That
is
they
don't
care
who
you
are
all
they
care
about?
Is
that
you've
put
enough
collateral
up
in
the
protocol
and
that
if
the,
if
certain
conditions
get
met,
the
protocol
has
a
way
to
liquidate
and
make
sure
that
it
stays
whole
on
the
loan.
So
that
is
the
mechanic
of
liquidation,
how
it
actually
sells
that
the
assets
who
buys
them
those
details
vary
and
and
at
agora,
we're
working
on
mechanisms
there
as
well,
but
at
a
high
level.
This
generally
works
the
same
for
all
the
over-collateralized
lending
protocols.
A
Okay-
and
so
I
showed
before
for
basic
attention
token-
there
was
a
massive
difference
between
the
rate
that
was
supplied,
the
the
rate
earned
by
suppliers
and
then
the
rate
that
it
cost
to
borrow,
and
the
question
is:
how
does
that
rate
get
calculated?
And
again?
You
know
this
is
up
to
the
protocol,
but
both
aveda
and
compound
and
and
other
lending
protocols
do
this.
A
Similarly,
which
is
the
the
protocol,
has
a
certain
amount
supplied
for
a
given
asset
and
then,
let's,
let's
just
stick
with
ethereum
as
as
that
asset,
and
when
there
is
not
a
lot
of
borrow
demand,
the
rate
is
lower
and
they
just
have
a
function.
That
basically
says
what
is
the
utilization
rate
of
of
that
pro
of
that
asset,
so
I
use
borrow
demand
just
now,
but
but
really
its
utilization
rate
is
the
rate
of
borrows
versus
the
rate
of
supply.
A
So
when
utilization
is
really
low,
they
want
to
stimulate
more
borrowing,
so
the
rate
is
lower,
but
as
utilization
approaches
100,
it
typically
hits
a
sort
of
a
inflection
point
where
suddenly
the
rate
gets
higher
much
faster,
and
that
is
to
incentivize
people
to
pay
back
the
debt
that
they
have
they
have
taken
out,
because
the
protocol
needs
to
make
sure
that
it
has
enough
for
people
to
withdraw
right.
A
So
if
I
have
supplied
ethereum
to
a
lending
protocol,
I
expect
to
be
able
to
withdraw
that
from
the
pool
whenever
I
want
to,
and
to
be
able
to
do
that.
The
protocol
needs
to
have
ether
available
to
me
and
therefore
there
needs
to
be
enough
waiting
sitting
around
waiting
and
it
incentivizes
that
by
making
the
rate
higher
so
again,
these
are.
These
are
just
the
mechanics.
This
is.
A
This
is
how
this
works
to
allow
it
to
fluctuate
without
needing
human
intervention
right,
so
the
utilization
rate
determines
how
how
that
happens,
and
these
these
curves
are
different
in
most
of
these
protocols.
These
pro
these
curves
are
different
depending
on
the
asset.
A
Okay,
so
I
I
said
that
one
of
the
goals
here
was
to
talk
about
terms
that
you
see
in
the
industry
and
one
that
has
been
seen.
A
lot
is
yield
farming
and
really
you
know
it
just
comes
from
the
idea
that
I
I've
mentioned
two
lending
protocols.
You
can
imagine,
there's
a
whole
bunch
of
additional
protocols
doing
similar
things.
A
They
may
have
different
rates
that
you
could
earn
for
your
ethereum
or
your
usdc
or
or
whatever
asset
it
might
be,
and
the
term
yield
farming
just
came
from
the
idea
that,
as
a
user,
I
might
want
to
move
my
my
assets
quickly,
depending
on
who's
paying
me
the
best
rate
who's
paying
me
the
best
yield-
and
you
know
this
this
metaphor
got
stretched
a
whole
lot.
People
started
talking
about
rotating
crops,
I'm
just
a
humble
farmer.
A
You
also
may
think
well
what,
if
I
just
built
something
that
would
automate
that
for
for
for
people
right
for
myself
and
for
other
people,
and
that
is
exactly
what
year
in
finance
was
so
urine
finance
launched
in
this
last
summer.
I
think
somewhere
around
july
and
is
now
over
a
billion
dollar
valuation,
largely
just
based
on
automating,
this
function
for
users
and
saving
on
gas,
so
there
have
been
others
as
well.
A
Okay.
So
this
is
the
end
of
the
lending
section,
but,
as
I
said,
this
is
really
just
scratching
the
surface,
if
you're,
if
you're
interested
in
lending
either
just
generally
or
if
you
want
to
build
a
product
here,
there's
a
few
things
that
you
should
look
into
so
first
I've,
I
skimmed
across
flash
loans.
A
It's
a
really
interesting
innovation
that
is
specific
to
to
synchronous
blockchains
like
ethereum,
I
didn't
talk
about
minting
versus
supply
models,
so
in
this,
in
these
examples
I've
been
talking
about
supply
models
where,
for
for
every
asset,
that's
borrowed,
it
had
to
be
supplied
by
somebody
else.
There
are
also
minting
models
of
of
lending,
which
is
more
like
maker
dao,
where
you
contribute
assets,
and
then
the
protocol
actually
mints
a
new
synthetic
asset
in
in
maker
dow's
case
it's
dye,
which
then
you
can
use
and
that
is
treated
as
a
debt
position.
A
So
a
lot
of
a
lot
of
the
mechanics
are
very
similar,
but
the
fact
that
it
doesn't
need
to
be
supplied
means
that
there
are
differences
there
and
that
that
again
is
the
model
that
agorik
will
use
with
some
variations
in
our
treasury
as
well.
A
Some
open
active
areas,
so
it
doesn't
mean
nobody's
working
on
these,
but
it
means
that
it
at
least
to
me
it
doesn't
seem
like
there
are
clear
leaders
yet
under
collateralized
lending,
so
figuring
out.
That
out
is
difficult
but
would
be
a
you
know,
huge
boon
to
the
d5
economy.
A
Right,
if
you
can,
if
you
can
actually
lend
people
lend
to
people
based
on
reputation
successfully,
that
would
supplant
a
whole
bunch
of
traditional
finance,
stable
rates,
so
ave
does
offer
stable
rates,
and
I
there
have
been
some
protocols
that
have
have
launched
with
that.
But
right
now
you
know.
I
showed
that
utilization
curve.
Those
rates
vary
right.
A
So
if
you
take
a
loan
out
and
you're
holding
it
for
several
months,
your
rate
will
change
during
that
time
and
it's
hard
to
know
what
you
will
pay
over
the
course
of
those
months
and
that's
that's
a
difficult
thing
for
users
so
figuring
out
how
to
manage
stable
rates
is
important
and
then
tranched
risk,
and
this
may
or
may
not
be
true.
A
You
know
a
true
lending
thing,
but
in
in
finance
you
often
see
bonds
get
listed
with
tranched
risks
so
that
you
know
one
one
group
might
take
all
the
risk
of
default
up
until
a
certain
point
and
they
are
the
riskiest
tier
and
they
earn
the
most
that
sort
of
thing
it
allows
you
to
take
one
one
stream
of
cash
flows
and
separate
it
into
multiple
different
products.
So
again,
people
are
working
in
these
areas,
but
it
still
is
open
and
it's
something
that
you
should.
A
If
it's
something
that's
interesting
to
you,
you
should
look
into
it
all
right.
I'm
gonna
pause
for
a
minute.
I
see
a
few
comments
coming
in
here.
If,
if
there
are
questions,
please
do
send
them
along
I'll,
see
I'll,
see
comments
in
youtube
and
you
know
please
do
stop
me
if
you've
got
any
questions.
Otherwise,
I'm
going
to
keep
rolling
on
automated
market
makers.
A
All
right:
let's
go
okay,
so
automated
market
makers
and
I'm
going
to
go
back
to
this
slide
after
I
talk
a
little
bit
about
them,
but
the
goal
here
really
is
just
to
show
you
the
it's
going
up
and
to
the
right
right.
So
there
are.
The
growth
in
automated
market
makers
has
been
significant
over
over
the
course
of
the
last
year,
and
this
is
just
volume
being
processed
by
the
primary
ones.
A
So
I'll
come
back
here
to
talk
about
individual
amms,
but
if
we
to
frame
what
an
automated
market
maker
is
doing,
it's
useful
to
think
about
it,
as
in
contrast
to
how
you're
used
to
looking
at
a
an
order
book
in
it.
You
know
if
you
buy
your
own
stocks
or
something
like
that,
so
the
usual
exchange
model
is
called
a
central
limit
order
book
and
the
way
that
works
is
you
know
each
line
here
is
a
specific
price
and
there
may
be
multiple
orders
that
people
put
in.
A
So
if
you're
familiar
with
putting
in
a
limit
order,
you
know
I
I
might
want
to
buy
atoms,
for
example,
and
if
it's
trading
at
twenty
dollars,
I
want
to
put
my
limit
order
in
at
19.50
or
something
like
that
with
the
idea
that
if
the
price
drops
a
little
bit
I'll
get
filled,
and
I
will
now
have
saved
a
little
bit
versus
buying
it
directly
at
20..
That
might
be
the
way
a
user
thinks
about
it.
A
Institutions
do
this
in
a
much
more
regimented
and
sophisticated
way,
but
but
an
order
book
is
just
saying
what
are
all
the
orders
to
buy
and
what
are
all
the
orders
to
sell
and
if
any
of
them
match
up,
I'm
going
to
execute
them,
and
that
is
so
and
what
that
means
is
that
every
time
you
are
buying
somebody
is
selling
and
the
the
exchange
is
just
matching
you
with
some
other
real
entity
at
that
time,
and
so
the
difference
with
an
automated
market
maker
is
that,
instead
of
one
buyer
and
one
seller,
what
you
have
is
you
have
sellers
effectively
called
liquidity
providers,
contributing
liquidity
to
a
pool,
and
then
you
have
buyers
which
may
come
in
at
different
times
and
and
are
able
to
buy
whenever
they
want
to.
A
So
we'll
go
into
a
little
bit
more
about
how
the
actual
mechanics
of
this
work.
But
it's
really
important.
The
real
innovation
for
automated
market
makers
was
separating
the
need
to
have
a
seller
for
every
buyer,
or
at
least
in
my
view,
that
that
truly
is
what
was
revolutionary
about
it.
A
So,
let's
walk
through
an
example
and
I'm
gonna,
I'm
gonna
try
to
introduce
concepts
sort
of
gradually,
but
this
is
specific
to
an
xy
curve.
Automated
market
maker
I'll
talk
a
little
bit
more
about
what
that
means.
So
I've
got
three
groups
here:
I've
got
a
liquidity
provider.
I've
got
the
pool
itself,
which
is
the
amm,
the
automated
market
maker,
and
then
I've
got
a
trader.
A
So
the
pool
starts
with
a
hundred
eighth
and
a
hundred
thousand
usdc.
This
is
assuming
a
thousand
dollars
per.
If
and
again
this
is
just
arbitrary.
You
know
the
pool
will
have
to
be
seated
by
somebody.
It
doesn't
just
get
assets
right,
but
in
this
example
this
is
where
it's
starting
and
then
you
have
a
concept
here
called
k
which
I'll
I'll
talk
about
a
little
bit,
but
for
now
what
you
know,
what
the
pool
knows
is
that
it
has
a
hundred
eighth
and
a
hundred
usdc
or
a
hundred
thousand
usdc.
A
Excuse
me:
now:
a
liquidity
provider
comes
along
and
say:
that's
you
or
me,
and
I've
got
one
each
and
a
thousand
usdc,
and
I
can
contribute
that
to
the
pool
and
what's
important
is
I
have
to
contribute
this
in
the
ratio
that
it
currently
exists
in
the
pool
so
right
now
there's
a
thousand
usdc
for
every
heath
in
here
hundred
thousand
to
a
hundred,
and
so
I
need
to
have
one
each
and
a
thousand
usdc.
A
If
I
wanna
put
in
you
know,
I
can
do
500,
usdc
and
half
an
eth,
but
the
ratio
has
to
be
equivalent.
So
I
contribute
it.
There's
now
101
each
in
the
pool
and
101
000
usdc
in
the
pool,
and
so
as
a
result,
I
have
0.99
percent
of
the
pool
you
can
see
here.
The
price
started
at
a
thousand
dollars
and
I'm
using
dollars
as
a
convenience.
Really
it's
just
a
thousand
usdc
per.
A
If
and
after
I've
contributed,
the
price
is
still
at
a
thousand
dollars,
but
now
a
trader
comes
along
and
the
trader
says
I
would
like
to
buy
five
east
from
this
pool
and
what
they
don't.
What
the
trader
doesn't
necessarily
know
is
what
the
prices
they
will
get.
The
pool
can
tell
them
at
the
time
they
say
I
would.
A
I
would
like
to
trade,
but
or
they
could
calculate
it
themselves
if
they
know
how
much
liquidity
is
in
the
pool,
but
for
this
purpose
let's
say
I
just
know
that
I
want
to
buy
five
eth
and
I'm
relatively
price
insensitive,
so
whatever
price
it
ends
up
at
that's
fine
with
me,
I
buy
5e
for
usdc
and
you
can
see
now
that
what
happens
after
that
is
the
eave
drops
by
five
and
the
usdc
increases
by
more
than
five
thousand,
and
the
reason
for
that
is
all
the
pool
knows.
A
The
reason
why
it's
called
an
xyk
pool
is
the
k
is
invariant
here.
So
right
now
k
is
10
million
201
201
thousand,
when
the
eth
changes
k
has
to
stay
the
same
and
as
a
result,
you
get
what
the
usdc
balance
needs
to
be
to
make
that
make
that
true,
so
k
can
increase
or
decrease
when
liquidity
is
added
or
removed,
and
you
see
that
here,
but
on
trades
k
stays
the
same,
and
that
is
what
determines
the
price.
A
So
the
elegance
of
this
is
that
the
pool
only
needs
to
know
how
much
each
it
has
how
much
usdc
it
has
and
what
k
is
or
excuse
me
and
and
the
curve
that
it
is
using
to
transform
them
and
k
falls
out
of
that.
So
that's
all
the
pool
really
knows
it
doesn't
know
anything
about
the
what
the
price
should
be.
It
doesn't
know,
it
doesn't
really
know
anything.
It
just
knows
I'm
holding
these
assets
and
I've
got
this
this
curve
to
transform
them.
A
But
what
falls
out
of
that
is
the
price,
because
now
there
is
a
certain
amount
of
usdc
and
a
certain
amount
of
eath
in
the
pool
and
the
price
now
is
at
a
thousand
one
hundred
and
six
dollars
and
88
cents.
Okay
and
I'm
seeing
a
question
come
in.
Does
the
user
have
an
option
in
trading
with
a
specific
pool
and,
and
the
answer
there
is,
it
depends
on
exactly
what
you
mean
by
the
question,
but
the
answer
is
yes
right.
A
So,
as
a
as
a
user
going
to
uni-swap,
for
example,
this
pool
is
ethan
usdc,
they
have
a.
They
have
separate
pools
for
all
the
different
asset
pairs
that
they
need
to
use
and
sort
of
so
sometimes
they
might
route
through
multiple
pools
so
like,
for
example,
if
I
want
to
trade
to
some
other
asset
that
they
don't
have
a
specific
pool
for
it
might
go
through
two
pools,
but
you
have
a
choice
of
where
you
want
to
trade.
A
The
curve,
though,
is
specific
to
the
protocol
so
or
at
least
in
most
cases,
so
in
in
uniswap's
case,
all
their
curves
are
xyk
curves
and
that
that
is
true
for
for
anything,
that
is,
that
is
done
in
that
protocol.
So
you
don't
necessarily
you
can't
necessarily
say
I
want
to
trade
these
same
assets
with
a
different
curve
that
has
to
already
exist,
and
the
liquidity
providers
have
to
agree
to
add
their
assets,
to
a
specific,
specific
pool
with
a
specific
curve.
A
Okay,
and
just
just
to
fill
out
this
this
example
here
in
this
case
the
user
now
wants
to
withdraw
their
liquidity
and
remember
they
have
0.99
percent
of
the
pool
they
withdraw
0.99
and
now
they're,
taking
out
something
different
than
what
they
put
in
0.95,
eth
and
1052
usdc,
and
that
is
because
trades
occurred.
The
price
changed,
and
so
their
ratio
has
changed
as
a
result
of
of
those
trades,
and
that
gives
way
to
the
idea
of
impermanent
loss,
which
we
will
talk
about.
A
Okay,
I'm
gonna
pause
here
for
any
additional
questions
as
before
I
go
forward
and
again,
this
is
a
bit
of
a
toy
example,
but
hopefully
this
shows
the
mechanics
of
how
an
amm
actually
works.
A
Okay,
so
in
permanent
loss,
probably
the
worst
named
concept
in
defy
and
that
that's
a
difficult,
difficult
competition,
but
it
has
confused
more
people
than
it
has
informed,
but
the
term
remains
so
I'm
going
to
explain
it
as
impermanent
loss,
and
I
apologize
for
that.
I
think
so
dean
our
ceo
likes
to
say
give
me
more
of
whichever
of
these
tokens
does
worse
and
that
that
largely
does
explain
impermanent
loss
right
so
more
technically
as
an
lp
as
a
liquidity
provider.
A
You
are
short
the
relative
volatility
between
the
tokens
you
provide
and
to
make
that
more
concrete.
In
this
case,
I
added
one
eth
and
1000
usdc
the.
If
price
went
up
as
a
result
of
trading-
and
you
know
here-
I'm
just
showing
one
trade,
but
there
could
be
thousands
hundreds
of
thousands
of
trades
before
I
remove
my
liquidity.
A
But
all
that's
important
is
that
the
result
of
those
trades
was
that
the
east
east
price
went
up
and
then,
when
I
withdraw,
I
actually
have
I'm
withdrawing
less
ef
and
more
usdc,
so
heath
price
went
up
and
I
get
less
of
it.
Usdc
versus
eth
went
down
and
I
get
more
of
it
and
that's
really
the
the
mechanics
of
the
liquidity
pool
right.
A
They,
you
are
withdrawing
the
same,
the
equal
value
in
usd
or
any
third
price
of
the
of
the
assets
that
you
put
in
and
really
what
that
means
is,
as
each
price
goes
up,
you're
you're,
providing
liquidity
for
our
traders,
and
so
traders
are
we
get
this
right.
As
the
east
price
goes
up,
traders
are,
they
are
buying
heath,
which
means
removing
each
from
the
pool
and
giving
you
the
other
token
usdc.
A
So
the
pool
weight
will
have
less
and
less
eath
and
more
and
more
of
the
other
token,
and
through
the
nature
of
you,
having
provided
the
sale
to
them
through
that
through
that
rise,
you
actually
lose
relative
to
holding
both
those
assets
alone,
and
that's
that's
really.
What
impermanent
loss
means
is
as
the
asset
prices
which,
when
you
give
them,
they
are
equal
because
they
they
have
to
be.
As
the
asset
prices
diverge,
the
more
they
diverge,
the
more
your
returns
will
lag
versus
just
holding
the
assets
themselves.
A
However,
the
counterpoint
is
that
you
earn
fees.
So
I
I
because
this
is
a
toy
example.
I
didn't
add
it
here,
but
in
this
trade
there
is
a
0.015
e
fee,
because
I
just
picked
unit
swap
uniswap
has
a
0.3
percent
fee
and
that
gets
added
to
the
pool
over
the
time
over
time
which,
as
an
lp,
you
get
a
share
of.
So
what
you
are
betting
when
you
are
providing
liquidity.
Is
that
the
losses
of
the
impermanent
losses?
A
So
hopefully
that's
clear
again.
Ask
questions
if,
if
this
isn't
but
impermanent
loss
really
just
means,
as
the
prices
diverge,
you
will
lose
versus
holding
okay.
So
I
I
talked
a
little
bit
about.
I
guess
I
haven't
gotten
there
yet
one
of
the
use
cases
of
amms
is
liquidity
mining,
and
this
is
again.
A
An
actual
exchange
listings,
especially
for
unknown
projects,
are
extremely
expensive
so
that
that's
a
huge
barrier
that
was
a
huge
barrier
to
launching
a
new
project,
but
now
because
of
uniswap,
because
there
are
and
sushi
and
other
amms,
because
there
are
these
permissionless
amms,
you
can
have
your
users
create
liquidity,
pools
for
you
and
the
way,
the
way
you
incentivize.
That
is
you,
give
away
your
token
to
people
that
are
doing
that.
A
So
in
many
cases
a
project
will
launch
and
they
will
distribute
their
token
in
in
part
or
entirely
just
by
people
bringing
their
liquidity
into
the
system.
So,
for
example,
they
a
lending
protocol
might
want
you
to
contribute
eth
so
that
eth
can
then
be
borrowed
out.
They
might
distribute
their
token
to
people
that
contribute
eth.
So
you
bring
in
your
eth
and
you
get
some
of
the
some
of
their
their
native
governance.
Token.
A
There
is
a
separate
pool.
That
really
is
what
liquidity
mining
is.
Where
you
get
that
governance
token,
and
then
you
supply
that
governance
token,
with,
for
example,
usdc
to
another
pool,
so
that
people
with
usdc
can
just
buy
your
token
directly
right.
So
the
way
to
get
the
token
to
begin
with,
is
you
supply
some
liquidity
and
some
other
token,
and
you
gradually
get
the
governance
token
of
the
project.
But
then
you
can
supply
that
governance
token
and
allow
other
people
to
buy
it,
and
that
is
the
liquidity
in
liquidity
mining.
A
So
projects
will
launch.
They
will
ask
you
to
please
supply
our
token
in
you
know,
in
a
pool
with
usdc
or
with
some
other
other
asset,
and
we
will
reward
you
with
additional
tokens
so
that
you
know
there
is
in
the
in
d5
parlance,
that's
pool
one
and
pool
two
and
a
liquidity
mining
campaign
to
be
successful
there.
Really.
A
What
you're
trying
to
do
is
you're
trying
to
bring
attention
to
the
project,
you're
seeding,
initial
liquidity,
so
that
you
know
that
that
pool
one
where
people
are
bringing
ethereum
in
so
that
it
can
then
be
lent
out.
Hopefully,
that's
doing
something
important
for
your
actual
product.
That's
an
important
part
of
this
some
projects,
some
projects
do
liquidity
mining
without
actually
having
a
value
for
liquidity,
which
is
a
problem.
Your
goal
is
to
also
broadly
distribute
the
tokens.
A
You
want
a
lot
of
people
owning
that
and
have
a
broad
community
base,
and
then
it
also
kick-starts
the
actual
price
discovery
for
your
token
itself
and
trading,
and
eventually
the
hope
is
that
the
liquidity
that
is
being
put
in
an
amm
that
you
are
incentivizing
starts
is:
is
there
sort
of
more
permanently,
even
after
your
incentives
end?
I
realize
that
I
am
significantly
over
time
here.
A
So
I'm
gonna,
I'm
gonna
roll
a
little
bit
faster
as
a
culture
moment
pool
two
farmers,
sometimes
that's
sometimes
known
as
the
death
pool.
A
If
you
are
new
to
d5,
just
know
that
by
entering
pool
two
the
which
is
the
governance
token
and
some
other
token,
you
are
exposing
yourself
to
a
pretty
significant
amount
of
risk,
something
something
important
to
know
if
you're
interested
in
taking
a
m's
further
look
at
different
curved
models,
so
stable
swap
was
released
by
curve,
and
that
now
is
one
of
the
largest
amms.
So
this
this
maroon
color
is
curve.
That
was
just
an
innovation
in
the
xyk.
You
know
versus
xyk.
A
They
did
something
different
and
again
no
time,
but
please
please
do
look
into
that
sort
of
thing.
That's
really
important.
People
are
working
on
capital,
efficiency
of
liquidity,
mitigating
and
permanent
loss.
If
they
can,
there
will
always
be
a
permanent
loss
in
amms,
but
there
are
potential
ways
that
you
can
either
compensate
users
for
it
or
attempt
to
help
mitigate
it
and
then
combining
amms
with
order
book
functionality.
A
All
right,
so,
let's
talk
a
little
bit
about
defy
and
agorik.
You
know
from
from
our
standpoint,
we
we
talk
a
lot
about
the
technical
properties
of
our
system,
the
security
properties.
Why
we
think
building
in
agorik
makes
a
lot
of
sense
for
those
reasons,
but
I
I
think
really
what
it
boils
down
to
is
just
being
able
to
build
faster
and
and
maintain
the
level
of
safety
you
need.
A
So
there
there's
a
whole
bunch
of
reasons
for
that,
but
the
frameworks
that
agora
provides
are
tailor
made
for
building
d5
products
right
that
we
are,
we
we
have
frameworks
for
assets,
asset
exchange
being
able
to
offload
the
exchange
of
assets
between
parties
to
a
trusted
protocol
level
thing
and
again
it
because
this
is
an
attack
on
a
gorick.
I'm
not
gonna,
I'm
not
going
to
go
deep
on
this,
but
really
what
this
all
adds
up
to
is
being
able
to
build
faster.
A
The
the
our
our
native
protocol
itself
has
lending
an
amm
built
in
so
as
an
agora,
token
holder.
You
are
helping
to
both
secure
the
network,
but
also
secure
the
economy,
and
because
of
that
there
there
will
be
components
that,
as
a
third-party
developer,
you
can
reuse
that
we
are
using
at
our
protocol
level.
A
So
you
can
be
sure
to
build
on
top
of
those
understand
the
frameworks
we're
using
and
there
will
be
a
lot
of
examples
for
you
as
a
builder
and
beyond
that,
it's
something
that
we
understand
as
a
team,
because
it
is
core
to
what
we're
doing
okay.
So
I
I
want
to.
I
want
to
summarize
quickly
thoughts
on
launching
your
project,
so
first
off
obviously
make
sure
that
you're
solving
a
real
problem.
A
A
Oh,
I'm
gonna
have
a
protocol
where
somebody
can
bring
in
assets
and
other
people
can
borrow
them
the
those
those
are
already
done
right,
they're,
they're
winners
in
those
markets
in
large
part,
so
you
have
to
understand
where
the
holes
are
what
they
are
not
doing.
Well
enough.
You
know
if
they
have
a
problem
with
their
liquidation
model.
Is
that
something
you
can
fix?
A
So
I
call
out
a
couple
recent
examples
here:
alchemist
launched
about
a
month
ago,
and
their
model
is
lending,
but
no
liquidations
or
you
know
it
part
of
your
loan
pays
off
the
other
part
and
and
again
it's
small
tweaks,
small
tweaks
to
the
way
these
protocols
function,
but
that
make
a
big
difference
to
users.
So
understanding
the
user
behavior
around
this
stuff
is
really
important.
A
Can
you
create
a
liquidity?
Flywheel
right?
Is
liquidity
actually
driving
something
important
for
your
protocol?
Can
you
have
liquidity
coming
in
driving,
better
experience
for
users
driving
more
liquidity,
driving,
better
experience,
token
distribution
is
a
part
of
your
product
offering
so
think
about
that.
Think
about
how
broadly
you
want
it
to
go.
Think
about
how
you
want
that
distribution
to
happen,
and
then
I
and
I
should
have
added
governance.
Community
building
and
governance
is
part
of
your
product
offering
too.
A
So
in
my
definition,
early
on,
I
talked
about
credibly,
neutral
protocols
and
the
reality
is
when
you
launch
nobody.
Nobody
has
credibly
neutral
means,
there's
no
governance
or
extremely
minimized
governance,
but
it
is
impossible
to
do
that
and
a
bad
idea
to
do
that
when
you
first
launch.
So
the
question
is:
what
is
your
pathway
to
get
to
that,
and
how
do
you
communicate
that
to
your
community
and
are
they
a
part
of
it?
Are
they
bought
in?
If
you're
going
to
have
decentralized
governance?
Do
you
have
people?
A
Do
you
have
people
as
part
of
your
community
that
will
do
a
good
job
with
it
right
that
community
building
is
truly
important
and
I
apologize
for
rushing,
but
I
just
want
to
make
sure
we're
good
on
time.
So
so,
finally,
just
summary
thoughts
d5,
it
provides
real
value
for
users.
There
are
a
lot
of
language
issues
and
concepts
that
are
very
specific
to
d5
that
make
it
seem
unapproachable,
but
it
when
you
really
kind
of
dig
in
it.
A
The
concepts
aren't
that
hard
right,
it's
it's
primarily
easy,
algebra
or
just
a
term
that
that
is
unnecessarily
specific,
and
so
it's
not
as
impenetrable
as
it
may
seem
and
as
a
reminder
just
succeeding
here
requires
increasing
specialization.
Now
you
need
to
know
your
market
better.
You
need
to
be
able
to
build
faster,
and
that's
that's
in
large
part.
A
So
with
that,
I
think
I'll
end,
but
I'm
happy
to
spend
a
minute
or
two
waiting
for
questions.
If
there
are.
A
A
All
right:
well,
I
really
appreciate
the
time.
Please
do
join
our
discord
if
you
haven't
already,
if
you
have
questions
here,
feel
free
to
reach
out
to
me
directly
or
any
of
us
in
discord
and
we're
looking
forward
to
seeing
you
soon.