►
From YouTube: Governance and Risk Meeting: Ep. 28
Description
The theme for this call was 'Demand/Supply Imbalance' and as such we focused on issues surrounding that. Steven spoke a bit about what Decentralized governance means to MKR holders. Cyrus reviewed the numbers since the Stability Fee raise and provided some analysis and solicited suggestions for next steps to manage the peg. We then spend the rest of the time in a general Q&A session.
Discussion Thread: https://www.reddit.com/r/mkrgov/
Podcast version: https://soundcloud.com/makerdao/sets/governance-and-risk
Transcripts (Delayed by 48 hours) : https://github.com/makerdao/community/tree/master/governance/transcripts
Governance and Risk Framework - Part 1: https://medium.com/makerdao/makerdao-governance-risk-framework-38625f514101
Governance and Risk Framework - Part 2: https://medium.com/makerdao/makerdao-governance-risk-framework-fc8939f3d4e9
Foundation Proposal: https://medium.com/makerdao/foundatio...
Governance and Risk Framework - Part 3: https://medium.com/makerdao/makerdao-governance-risk-framework-part-3-7a4c620f4077
Website: https://makerdao.com
Twitter: https://twitter.com/makerdao
Reddit: https://www.reddit.com/r/MakerDAO/
Chat: https://chat.makerdao.com/home
Email: info@makerdao.com
A
Hello,
everyone
welcome
to
the
March
28th
edition
of
the
scientific
governance
and
risk
meeting
it
also,
coincidentally,
is
our
28th
meeting.
So
that's
that's
a
lot
of
meetings.
We've
gone
over
a
lot,
so
I
think
that
we're
in
a
good
place,
though
so
I'm
I,
continue
to
look
forward
to
these
all
week.
I
think
there's
some
interesting
stuff
that
we're
going
to
be
able
to
tackle
today
as
well.
Preamble,
of
course,
is
that
we
are
very,
very,
very
interested
in
soliciting
as
much
information
as
we
can
from
the
community.
A
You
guys
are
all
enormous,
ly,
smart
and
plugged
in
people.
We
need
to
get
your
feedback.
So
if
you
have
a
question
or
you
have
a
comment-
don't
wait
for
an
opening
just
type
it
into
the
chat.
If
you
don't
have
a
microphone
or
interrupt,
if
you
do
have
a
microphone,
the
conversations
are
heating
up
a
bit.
A
very
energetic,
very
passionate,
that's
great.
A
Also
one
thing
I
want
to
do
is
that
if
you're
having
a
have
a
question,
you
won't
have
answered-
or
you
want
to
become
part
of
the
greater
discussion
and
you
asking
the
chats
will
make
an
effort
to
bring
that
into
the
main
conversation,
because
I
want
to
try
and
avoid
a
situation
where
we
have
in
furious.
The
interesting
debates
happening
in
the
chat
while
there's
up
here
is
the
interesting
debate
happening
in
the
same
call.
So
I
want
to
try
and
sink
those
things
up
so
we're
on
the
same
page.
B
Firstly,
welcome
everybody,
and
thank
you
very
much
for
for
joining.
This
call
also
an
expression
of
gratitude
with
so
many
folks
in
the
community
that
are
actually
contributing
towards
the
analysis,
contributing
data.
It
is
very
much
appreciated,
so
please
keep
that
coming
to
that
end.
I
just
want
to
reiterate
a
few
things.
The
first
is
the
themes
that
we
want
to
cover
in
these
in
his
meetings.
As
you
know,
the
first
one
is
the
demand
and
supply
the
cyclic
ation.
You
know
what
do
we
think
about
that?
How
do
we
measure
it?
B
And
you
know
what
are
the
considerations?
The
the
second
theme,
which
will
obviously
start
getting
into
either
today
or
even
a
later
time,
is
collateral
types,
the
collection
portfolio
and,
obviously
the
parameter,
adjustments
and
calculations
around
that,
and
then
the
third
theme
is,
is
you
know,
basically
exogenous
risks.
These
are
the
three
themes
that
I
think
we
should
continually
try
to
focus
on,
as
they
actually
do
cover
the
broad
spectrum
of
the
the
decentralized
risk
and
the
scientific
evidence
principles
now.
B
But
please
remember
this
is
a
decentralized
scientific
governance
and
this
is
a
discussion
about
the
stability
fee
which
are
basically
considered
and
facilitated
by
the
naked
eye
Foundation,
which
is
just
yet
another
community
member
and
more
so
a
non-voting
community
member
and
what
we
ultimately
want
to
do
is
put
these
things
forward
so
that
the
decision
can
be
made
by
the
nkr
governance
again,
also
keep
in
mind
about
continuous
approval
voting.
The
essence
of
this
is
that
you
always
need
to
be
aware
of.
What's
going
on,
so
please
be
involved
as
much
as
possible.
A
Thank
you,
Stephen.
We
have
we're
in
a
situation
where
we
we
could
talk
to
the
risk.
We
have
a
few
speeches
and
stuff,
but
I
want
to
make
sure
that
this
conversation
stays
as
targeted
as
it
possibly
can,
while
keeping
everybody
in
the
callin
opportunity
to
sound
off
a
bit
more
than
we
had
in
the
last
few
calls,
because
there's
been
a
lot
to
go
over
so
I'm
not
going
to
monopolize
by
over
ten
minutes
of
the
meeting.
A
I
want
to
hand
it
off
to
Cyrus
almost
immediately
to
give
us
the
update
from
the
risk
team
and
have
him
talk
about
the
state
of
the
peg
and
anything
else
of
interest,
and
then
I
want
to
start
digging
into
some
of
those
things.
He
also
has
a
special
presentation
for
us
today,
which
I'm
sure
most
people
here
will
be
very
interested
to
see.
You
know
Cyrus
I
want
to
take.
C
C
All
right
so
in
terms
of
the
dye
price,
obviously
we're
not
still
quite
at
$1
I
mean
in
some
respects
not
even
close,
we're
just
still
hovering
in
the
97.5
to
low
98s
range.
Pretty
consistent,
I
will
I
will
say
that
one
thing
that
we
noticed
this
week
was
a
less
variance,
although
not
technically,
at
the
price
that
you'd
like
to
see,
you
saw
a
lot
less
volatility,
a
lot,
less
kind
of
trader
panic
or
whatever
or
speculation.
You
might
want
to
call
it,
and
you
know
this
is
not.
C
The
surprising
is
actually
something
that
at
least
slightly.
We
are
happy
to
see
in
the
sense
that
we
obviously
we
raised
that
stability
fee
to
seven
and
a
half
percent,
and
it
definitely
had
impacts
in
other
areas.
I'll
flip
back
to
this
chart
in
a
moment,
but
main
thing
that
we
saw
this
week
but
I
want
to
highlight,
are
these
pink
bars
over
here,
which
is?
We
saw
some
significant?
C
C
C
Yeah
I
mean
seven
million,
is
a
fairly
large
amount
of
guy
that
was
was
wiped
down.
A
very
informal
chain
analysis
revealed
that
you
know
if
I,
if
I
had
to
guess
it
was
basically
market
making
activities
from
you
know
box,
or
you
know
whomever
that
we
may
be
taking
advantage
of
kind
of
cheap
capital
to
deploy
strategies
at
seven
and
a
half
percent
stability
fee
not
entirely
likely
that
they
were
able
to
turn
a
profit
in
their
estimation.
So
we
saw
you
know
a
few
very
large.
C
Cdp's
close,
you
know,
I
think
about
a
million
million
each
so
yeah.
So
what
does
this
mean
for
the
dye
price?
You
know
the
dye
price
is
obviously
not
at
a
dollar
a
my.
As
far
as
I'm
aware
general
market
maker
inventories
are
roughly
the
same.
I'll
have
Joe
chiming
and
a
bid
on
that
and
whoever
else,
but
you
know,
I,
think
it's
important
to
note
that
there
was
still
a
positive
impact
in
other
senses,
namely
that
you
know
the
liquidity
that
you
might
have
expected
to
stand
in.
C
C
C
So
yeah,
you
know
this
is
more
of
a
long
term
long
term
picture
kind
of
over
the
last
month.
I
mean
it's
not
entirely
clear
from
this
chart,
but
we
definitely
saw
a
significantly
lesser
volatility
in
the
dye
price.
A
lot
in
the
last
week
compared
to
what
we'd
seen
early
on
in
March
and
that's
it
for
now,
I'm
gonna
pass
it
to
Joe.
If
you
want
to
say
anything
or
any
of
the
other
market
makers,
thanks.
D
Iris,
this
is
what
yeah
I'm
a
just
harp
on
one
thing:
the
volume
and
volatility
are
down
tremendously
and
I
think
sorry,
100
percent
right
on
that
have
seen
some
solid.
You
know
when
we
move
to
seven
half
percent
saw
a
lot
of
market
makers
come
in
with
solid
bids
and
felt
more
confident
in
the
bidding
like
98
cents,
98
and
half
cents.
This
is
over-the-counter,
but
still
a
lot
of,
especially
on
the
exchange
exchanges.
These
decentralized
exchanges
they're
still
not
the
volumes
falling
up
up
a
clef.
D
A
C
Mean
maybe
maybe
orthogonal
II
in
the
sense
that
I
mean
the
dye
market
makers
aren't
impacting
overall
crypto
ball
in
one
way
or
another,
but
in
terms
of
which,
which
exchanges
we
saw
the
dye
being
pulled
from
you
know,
I
wouldn't
be
surprised
if
we
saw
less
dye
volume
on
those
particular
exchanges.
Does
that
make
sense,
yeah.
E
I
think
that
not
enough
selling
is
being
done
on
the
idea
that
holding
die
and
earning
these
incredibly
good
interest
rates
is
a
very
lucrative
thing
to
do
in
a
much
much
less
risky
way.
So
right
now,
I'm
just
going
to
throw
out
some
stuff,
you
know,
Dharma
Lieber
is
offering
6%
compound
finance
is
offering
3
4
%
Expo
trading.
They
were
keeping
a
lot
of
their
their
interest
is
secret,
but
they
can't
find
enough
lender.
So
they
can't
find
out
people
that
are
offering
died.
They're
gonna
raise
their
interest
rate
for
the
next
contract.
E
E
Locked
in
locked
in
for
28
days
and
there's
another
company
that
just
came
out
which
I
forget
their
name
but
they're,
doing
exactly
what
dy/dx
is
doing.
Someone
wrote
about
it
in
the
maker
chat,
so
I
feel
like
there
needs
to
be
an
entire
kind
of
like
speculation
page,
but
for
people
like
myself,
who
are
trying
to
find
the
best
interest
rate,
and
you
are
trying
to
capitalize
on
the
fact
that
interest
rates
are
going
up
so
I'll
throw
up
the
link.
Forex
go
right
now
in
the
chat.
It's
Trading
dot-com,
backslash
land.
What's.
A
A
Mateo
has
been
writing
some
excellent
articles
and
so
I'm
not
sure
whether,
as
a
journalist,
you're
comfortable
with
being
a
part
of
the
conversation
that
you're
reporting
on,
but
you
mentioned
in
your
last
article
on
the
block
about
I'd
die
for
which
I
think
kind
of
aligns
with
one
of
the
things
that
would
be
this
talking
about
Patera.
Do
you
want
to
tell
us
about
that?
I
better,
sit
this
one.
Oh
sure,.
D
Can
you
guys
hear
me
alright,
yep,
yep
great
so
yeah
dipole
is
this?
Is
this
decentralized
inter
protocol
offer
great
it's
a
kind
of
grip
of
LIBOR,
and
the
idea
is
that
you
know
on
these
government
schools.
You
can
actually
use
dipole
as
some
kind
of
reference
metric
and
assess
the
the
effect
that
a
stability
fee,
hike
or
decrease
has
had
on
the
wider
dye
economy.
D
So
if
dipole
is
relatively
low,
the
borrow
rates
a
relatively
low
and
that
might
signify
that
the
stability
feed
needs
to
be
height
because
there's
still
a
lot
of
supply
on
the
market
and,
conversely,
once
typo
starts
to
rise,
you
you
can
tell
that
the
stability
P
has
has
had
its
effect.
Possibly
it
might
be
worth
then
reducing
the
stability
thing.
Yeah.
A
I
think
that's
an
amazing
concept,
and
probably
amazed
because
anybody
isn't
familiar
with
traditional
finance
would
not
be
amazed,
but
from
come
from
their
technical
space
yeah.
This
seems
like
a
great
idea.
We
should
do
it.
Everybody
else
has
been
doing
for
the
last
hundred
years
or
so,
but
yeah
and.
D
You
know
one
of
the
problems
right
now
with
using
dipose
just
that
he's
he's
finding
more
consistent
liquid
and
to
efficient
over
time
as
we
as
we
start
to
see
more
liquidity
and
more
anchor
anchor
the
quiddity
of
crunchy
plant
bones.
It
could
be
a
pretty
powerful
metric,
you
know,
and
on
a
metric
that
would
a
company
is
something
like
market
maker
inventor
as
well.
You
just
help
us
make
these
kinds
of
informed
decisions.
A
Yeah
I
love
it.
If
anybody
else
has
new
comments
about
the
type
or
idea
speak
up
down
or
we
can
move
on
to
some
of
the
other
questions,
but
I
really
want
to
start
seeing
that
reference
ring.
I.
Think
that
if
we
could
be,
if
we
could
map
that
back
and
overlay,
that
onto
one
of
these
charts
that
were
looking
at
for
stability
fee
impacts,
we
might
see
some
interesting
correlations.
A
Obviously,
the
crypto
space
is
not
rational
and
there's
there's
a
delay
between
cause
and
effect
that
I
still
haven't
wrapped
my
head
around,
but
it
would
be
very,
very
interesting
to
see
those
two
lines
compete
with
each
other,
all
right,
miss
Ashley.
Do
you
have
a
cross
that
you
wanted
to
share
with
us
again
today?
Oh
I
think
we
just
lost
them
all
right
I
might
have
missed
some
of
the
other
interesting
questions
in
the
chat.
I'm
scanning
them
right
now.
If
I
have
speak
up
otherwise
Syrus,
maybe
you
can.
C
Sure
it's
gonna
be
I,
mean
shifting
gears
for
a
bit.
So
what
happy
to
shift
that
was
just
kind
of
run
through
this
and
then
with
the
remaining
time
we
can
talk
about
whatever.
So
basically,
as
we
approach,
I
mean
there's
more
to
make
her
down
than
just
the
monetary
policy
right.
There's
this
huge
section
of
collateral
risk,
which
is
actually
for
those
of
you,
who've
been
around
for
a
while.
C
It
was
kind
of
the
primary
focus
of
this
call
week
in
and
week
out
up
until
a
few
months
ago,
when
you
know
the
the
guy
peg
started
to
kind
of
take
precedence,
it's
a
more
pressing
manner,
so
I
think
going
forward.
We're
gonna
try
to
maybe
either
divide
these
calls
until
to
two
parts
or
maybe
just
devote
a
little
bit
of
time
to
talking
about
each
section
over
time.
C
C
Okay,
so
this
is
kind
of
just
a
mental
mental
model
of
the
maker
Dow
risk
function.
Some
of
you
may
not
be
familiar
with
this
term.
The
decentralized
risk
function,
so
this
kind
of
embodies
how
we
are
approaching
risk
as
a
Dow,
and
the
key
word
here
is
obviously
decentralized,
and
what
does
that
mean?
How
do
we
achieve
it?
How
do
we
go
about
it
great
question,
so
the
goal
of
the
DRF
is
to
have
kind
of
various
multiple,
diverse
risk
teams
that
are
working
cooperatively,
but
also
competitively
with
each
other
and.
C
So
I'm
gonna
jump
around
and-
and
this
offers
a
few
benefits
one
is
you
know
the
main
one
in
my
view
is
specialization.
I
think
it's
not
feasible
that
one
risk
team
can
manage
the
entirety
of
the
system.
If
you
can
imagine
the
kind
of
work
that
goes
into
deciding
the
monetary
policy
dealing
with
you
know
various
collateral
assets
managing
risk
parameters,
kind
of
maintenance
in
the
system
overall,
not
really
feasible-
that
one
team
can
kind
of
handle
everything
so
I
think
the
long
term
goal
is
to
see
specialized
risk
teams.
C
C
C
These
centralized
risk
teams
will
obviously
kind
of
ensure
that,
hopefully,
someone
is
always
working.
Someone
is
always
watching
over
the
system
and
working
towards
managing
risk
and
I
think
there
will
be
somewhat
of
a
competitive
environment
in
that
there
will
be
some
sort
of
reputation
or
status
to
be
earned
by
being
the
risk
team
that
will
ultimately
be
chosen
for
submitting
risk
parameters.
So
I
imagine
that
there
will
be
some
compensatory
advantages
of
being
a
good
being
a
good
risk
manager.
C
My
mind,
the
main
one
is
the
collateral
portfolio
which
comprises
of
a
qualitative
framework
where
you
dig
into
the
asset,
its
quality,
the
team
organization,
the
technology,
everything
that
has
to
do
with
the
qualitative
side
of
things,
and
then
there
is
quantitative
aspect
to
it.
Where
we'll
be
looking
at.
Basically
the
trading
profile,
liquidity
exchanges,
things
that
factor
into
how
we
would
determine
its
suitability
for
collateral
auctions.
Obviously
you
know
the
liquid
assets,
for
example,
would
have
to
be
dealt
with
differently.
C
Second,
big
class
is
monetary
policy,
which
you
know
we've
been
doing
for
the
past
few
months
now.
This
is
very
data,
analytics
heavy
teams
that
are
proficient
and
statistics
counter,
metrics
all
that
kind
of
stuff,
and
then
finally,
we
have
kind
of
a
class
of
you
know.
Technical
risks
in
relation
to
Oracle's
options
integrations
that
kind
of
stuff
interesting.
You
know
interesting
thing
that
a
lot
of
people
don't
think
about
for
auctions.
For
example,
is
you
know
the
length
of
these
auctions?
C
You
know
right
now
for
single
collateral.
We
just
have
the
three
percent
the
three
percent
marked
down,
but
obviously
for
multi
collateral,
we'll
be
shifting
to
an
option
structure.
You
know
I'm
sure,
there's,
there's
very
deep
game.
Theoretic
considerations
behind
auctions-
and
you
know
different
assets
with
different
liquidity
profiles-
would
require
different
different
structures.
So
you
know
I
could
imagine
just
a
ton
of
risk
work
being
done
just
purely
on
auction
structure.
C
C
You
know
basically,
for
anyone
who
spent
any
significant
amount
of
time.
You
know
investing
or
trading
in
the
space.
You
know
it's
just
a
never-ending
grind
to
you
know,
figure
out.
What's
really
going
on
behind
a
lot
of
these
organizations
resolve
always
some
interesting
edge
cases
when
it
comes
to
examining
the
crypto
space.
C
So
I
mean
eventually
we're
gonna
have
to
tackle
the
notion
of
how
we're
gonna
scale
just
adding
parameters.
You
know
if
we
have
hundreds
or
thousands
of
collateral
types
being
added,
there's
an
in
meat
to
be
some
sort
of
some
sort
of
cue,
some
sort
of
filtering
mechanism
where
we
are
able
to
kind
of
prioritize
which
collateral
types
we
want
to
see
and
asap.
C
So
you
know
qualitative
framework
can
kind
of
help,
filter
and
finally,
I
mean.
Obviously
the
maker
Dow
protocol
does
not
recognize
any
any
notion
of
qualitative
metrics,
so
we
are
going
to
have
to
score
these
collateral
types
into.
You
know
numerical
data
that
can
kind
of
wiggle
their
way
into
the
stability
fee.
C
C
C
C
C
B
B
That
looks
after
the
compensation
of
these
wrist
booms
and
the
other
thing
we
need
to
keep
in
mind
as
well
is
that
there
are
risk
teams
that
support
and
their
risk
teams
that
submit.
So
that's
also
another.
You
know
categorization.
We
need
to
look
at,
but
I
believe
if
we
just
start
from
the
the
first
idea
of
collecting
like
what
we're
doing
now,
the
necessary
contributions
from
from
community
members,
we
effectively
will
start.
B
You
know
slowly
growing
what
is
effectively
first
decentralized
function,
which
then
gets
split
up
into
the
teams
and
from
there
each
one
of
those
teams
can
give
a
good
sense
of
what
that
the
costing
would
look
like
for
themselves.
So
it's
kind
of
a
bespoke
thing
that
eventually
becomes
almost
like
a
a
little
markets
unto
itself
I've.
A
Been
trying
to
imagine
what
the
future
would
look
like
as
well,
which
is
a
bad
idea
in
this
crypto
space,
but
would
it
be
sort
of
a
combination
between
a
team
comes
to
us
and
says,
hey?
Look.
We
want
to
write
this
for
what
we
have.
These
chops
we've
done
it
before,
and
this
is
no
one's
going
to
cost
us
to
provide
you
with
this
data.
A
So
it
would
be
like
sort
of
a
one-off
kind
of
thing,
or
it's
going
to
be
more
of
a
subscription
type
of
deal
or
is
it
I
guess
it
doesn't
have
to
be
binary
right,
so
it
may
be
people
looking
sales
models
or
they
could
offer
us
a
monthly
subscription
off
the
foundation,
a
monthly
subscription
or
even
act
as
sort
of
like
a
separate
business
unit.
That's
just
constantly
feeding
us
data
I
mean.
B
You
kind
of
have
like
the
freemium
model,
where
you've
got
as
an
example.
A
regime
that
you
know
specifically
is
pretty
good
at
collecting
data
and
reproducing
that
data
collection
and
data
cleaning.
Very
clearly
to
the
point
that
everyone
you
know
in
make
a
governance
decides.
You
know
these
guys
do
a
great
job,
we'd
like
to
include
them
as
a
risk
team
and
the
data
that
they
provide
for
the
other
risk
team
members
can
also
be
affected
by
them.
So
in
the
beginning,
I
can
imagine
them
saying.
Well,
we've
done
this
work
here.
B
It
is
for
you
it's
either
something
that
is
just
a
small
subsection
of
what
they
do
in
general,
or
it
is
something
they
have
clips
and
focus
their
entire
team
on,
and
given
that
kind
of
scenario
you're
looking
at
it
and
saying
well,
what
did
it
cost
you
to
do
this?
You
know
what
is
it
that
you
want
to
be
compensated
for?
B
How
is
it
that
you
are
appropriately
motivated
by
the
profit
incentive
to
actually
contribute
appropriately,
and
they
could
also
keep
in
mind
that,
because
we
are
in
this
bootstrapping
phase
that
there
is,
you
know
what
80
million
died
in
circulation
right
now,
it's
not
going
to
be
a
complete,
economical
business
model,
but
it's
one.
That's
gonna
have
to
bootstrap,
along
with
the
system,
so
being
here.
First
getting
a
good
sense
of.
A
Others,
together
to
throw
another
spanner
in
the
works
that
there
seems
like
there's
this
room
here
for
an
organization
that
has
significant
maker
Holdings
would
be
doing
risk
type
functions
just
simply
to
protect
their
own
investments
if
they,
because
it's
our
own
best
interest
to
make
sure
that
the
community
has
the
most
the
clearest
and
most
accurate
information.
So
it
would
be
sort
of
a
pro
bono
thing
at
that
point.
Right!
Look
it's
just!
What's.
B
What's
interesting,
here
is,
if
you
take
a
couple
of
steps
back,
you're
gonna
quickly
realize
that
a
decentralized
risk
function
is
going
to
create
pretty
much
it's
phone
sort
of
economic
activity,
its
own
economic
world.
This
thing
way
that
you
know
governance
can
create
its
own
economic
world
as
well,
and
as
as
to
the
the
CDP
users
and
dye
users.
We
keep
forgetting
the
fact
that
no,
this
economic
growth
is
not
just
something
that
happens
on
chain,
but
actually
happens
on
each
side
of
all.
A
C
C
Even
you
know,
the
theorem
community
has
a
separate,
has
several
separate
chat
rooms,
so
they
just
solely
are
dedicated
to
research
conversations.
I
mean
unless
there's
some
something
I'm
missing.
I
think
it's
great
to
have
like
a
separate
forum
where
people
can
discuss
make
for
risk.
Specifically
that
doesn't
get
kind
of
clogged
up
with
everything
else.
That's
going
on
with
the
project.
A
A
A
C
And
I
imagine
I
mean
these.
These
risk
calls
that
these
weekly
words
called
would
be
a
place
for
us
to
kind
of
congregate
and
and
aggregate
the
research
that's
been
done
and
kind
of
critique
them
and
review
them
objectively,
I
mean
if,
if
we,
if
that
process
fails
and
sure
but
I
mean
we
have
a
pretty
pretty
clear
mandate
to
be
these
very
objectively
there's.
B
Two
aspects
to
this
that
we
need
to
keep
in
mind.
The
first
is
that
we
are
using
risk
tooling,
that
is
in
a
traditional
space
and
in
a
traditional
space.
The
breadth
and
depth
of
it
is
massive.
So
it's
it's
not
like.
You
can
come
out
and
decide
that
we're
going
to
use
the
astrological
process
of
the
stars
to
dictate
what
collateral
we're
going
to
use,
even
though
it
might
be
a
wonderful
idea
and
I
know,
there's
some
traders
that
do
that.
B
Unfortunately,
we
won't
fit
into
that
that
sends
the
scientific
rigor
that
has
been
formulated
appropriately
in
the
traditional
space
from
which
we're
borrowing.
So
there
is
going
to
be
a
very
clearly
defined
necessity
for
not
only
rigor
but
reproducible
research
as
well.
That's
that's
something.
I
need
to
emphasize
to
be
able
to
say
this
is
the
model
and
these
the
parameters
we
get.
Trust
me
is
not
enough.
It's
more
case
if
these
are
the
models
and
making
it
publicly
available.
B
That's
where
I've
got
the
data
from
this
is:
how
did
my
calculation
go
ahead
and
do
the
same
thing?
You
should
get
exactly
the
same
figure
and
that
allows
us
to
sidestep
folks
who
want
to
scream
the
loudest
about
how
good
their
models
are,
or
it
also
sidesteps
the
issues
that
you
might
have
folks
that
have
very
complicated
models
and
they
very
hard
to
digest
and
it
very
hard
to
to
reproduce,
but
they
may
be
entirely
correct.
B
We
have
to
also
make
sure
that
there's
a
tractability
there
that
is
understood
clearly
in
terms
it's
practical
using
a
system,
so
there's
various
aspects
of
this.
That
is
very
open
to
a
critical
debate
in
terms
of
its
usefulness,
and
this
is
specially
benchmarked
against
the
traditional
world,
so
we're
not
coming
in
from
the
cold
sort
of
just
you
know,
devised
and
all
kinds
of
new
methods
we
can
actually
use
that
which
is
really
available.
A
I'm
going
to
move
on
to
the
next
question
so
Jordan
asks:
will
there
be
a
risk
team
to
make
sure
all
the
collateral
pools
are
playing
nicely
together
like
could
the
digits
focus
risk
tune,
say
we
think
we
should
do
X,
but
then
the
foundation
says
I
get
that's
the
best
thing
for
you,
but
we
need
your
team
to
push
the
peg
up
so
that
this.
This
is
an
interesting
question
about
how
does
the
portfolio
interact
with
each
with
itself
and
other
members
of
the
of
the
collateral
walls
so
Cyrus
or
Steven?
C
I'll
take
a
stab
at
it.
Yeah
I
think
managing
the
overall
collateral
pool
is
a
sufficiently
complex
task
that
it
requires
its
own
risk
team.
So,
in
a
certain
sense
you
know
they
presented
the
jig's
focus.
Pristine
would,
you
know,
do
a
very
deep
dive
into
organization
and
and
submit
parameters
based
on
its
qualitative
specifications,
but
you
know
there
has
to
be.
C
There
has
to
be
another
risk
team
that
is
managing
the
overall
pool,
making
sure
that
you
know
everything
plays
nicely
together
and
that
team
cannot
also
be
doing
a
deep
dive
into
every
single
organization
in
the
clerical
I
mean
if
it
gets
into
into
the
high
numbers
so
yeah.
My
my
opinion
is
that
yeah
in
fact,
arguably
they're
the
most
important
risk
team
in
the
sense
that
they're
they're
managing
the
cross
correlations
between
different
asset
classes,
I
mean
dgx,
focus
risk
team.
A
A
Think
that's
the
part
that
I
keep
on
coming
back
to
is
even
managing
a
portfolio
that
you
have
absolute
control
over
the
inputs
and
the
outputs
is
a
is
a
non-trivial
exercise,
but
in
this
model
those
inputs
and
outputs
aren't
necessarily
under
the
kind
of
or
the
risk
teams
control.
So
it's
going
to
be
fascinating
to
watch
and
I.
Think
it's
going
to
be
one
of
the
primary
focuses
of
a
great
number
of
people.
It's
just
to
maintain.
They
have
the
assets
themselves
behind
the
scenes.
Can't
wait
to
see
this
thing.
A
Developed
I
want
to
open
up
the
discussion
here,
because
we
we
have
a
tendency
to
do
a
huge
data
dump
in
these
things
that
we
have
a
very
targeted
discussion
over
turn
over
a
rock.
There
is
very
interested
in,
but
we
don't
get
deeply
into
a
general
Q&A
session,
so
I
wanted
I
want
to
give
that
a
shot
today
and
unfortunately,
sometimes
when
you
try
to
push
these
things,
it
has
a
tendency
to
be
met
with
silence.
A
But
I
wanted
to
throw
a
few
different
topics
we
could
potentially
discuss,
and
if
anybody
has
any
comments
or
questions,
please
fire
them
out.
There.
We've
been
talking
about
debt
ceilings
on
and
off
for
the
last
couple
weeks
that
kind
of
got
lost
in
the
background
noise
of
the
stability
fee
increases,
and
it's
also
due
to
the
recent
stability
fee,
increase,
obviously
less
of
a
pressing
question,
but
if
people
are
still
interested
in
that,
I
would
love
to
talk
about
it.
A
Another
example
things
we
can
talk
about
is
the
way
that
the
governance
process
works
itself,
because
we
have
this
pins
that
we've
been
working
along.
Where
there's
a
governance
call
we
sort
of
come
to
some
general
conclusion,
and
then
we
we
have
that
just
continue
that
discussion
at
reddit
and
then
that
developed
it
into
a
governance
poll
historically
developed
into
a
governance
ball,
which
leads
into
an
executive
votes.
Our
governments
poll,
which
leads
to
the
next
call
where
we
discuss
the
results
of
that
on
that
Friday.
A
We
have
an
executive
and
I'd,
be
very
interested
to
hear
people's
opinions
about
how
well
or
poorly
that
system
is
working.
The
other
thing
that
I
want
to
discuss
in
these
calls
to
you
is
what
kind
of
tool
link
does
the
community
think
that
we're
missing,
or
they
would
benefit
from
and
another
sort
of
some
conversation
stuff?
Some
subjects
from
these
calls
is:
how
do
we
deal
with
voter
turnouts
and
voter
apathy
and
is
voter
apathy
even
a
problem,
there's
some
contention
about
that
which
I
think
is
worthy
of
some
discussion.
A
D
Well,
I
guess
one
of
the
questions
we've
brought
up
before
the
call
and
again
we
can
always
know
what
the
supply
number
is
but
and
I
think
it's
just
a
general
metric
on
every
call.
Even
if
we
estimate
it,
we
need
to
just
have
what
we
and
even
throwing
out
a
number
thinking
that
the
demand
is
X
and
if
obviously,
the
more
supporting
data
behind
it,
the
better.
But
so
much
of
our
how
we
get
to
the
peg
is
derived
around
the
equilibria
of
getting
supply
and
demand
more
or
less
to
match.
D
A
Yeah
I,
agree
and
I
think
that
the
chain
knows
this
thing
will
get
us
a
certain
distance
towards
that
goal
and
vishesh
and
Matteo
I
think
are
on
the
second
part
of
that
puzzle.
At
least
this.
The
only
second
piece
that
I
can
see
right
now
is
that
if
we
understand
what
the
dipole
is
and
we
have
a
clearer
understanding
of
usage
of
the
other
lending
platforms.
A
So
if
we
can
see
that
you
know
the
stability
fee,
change
occurred
and
compound
was
effectively,
and
you
know
beaver
and
whatever
the
new
one
that
we
all
just
discovered
EXO
something
and
if
they
were
all
affected
in
the
way,
that's
predictable
than
I.
Think
that
we
might
be
something
would
help
us
understand.
D
I
mean
at
the
end
of
the
day,
I
mean
it's
really
either
to
get
them
in
equally,
really,
they
need
to
increase
the
demand
or
decrease
the
supply
or
both
or
one
more.
Both
of
them
are
declining,
but
one
is
declining
more
than
the
other
or
both
are
accelerating,
and
one
is
accelerating
more
than
the
other,
but
until
we
at
least
throw
out
a
number
and
quantify
it
and
then
have
support,
which
will
be
a
work
in
progress
for
the
next
year
or
years
to
come
to
that
conclusion.
D
A
And
I
think
that's
the
question
we
understand
supply
because
we
know
exactly
where
it
comes
from.
It's
the
demand
part
that
that's
still
a
question
mark
until
we
understand
what
the
demand
is
is
the
demand
we
bring
up.
Is
that
demand
a
collateral
type
for
other
lending
platforms
and
then
what
percentage
of
that
demand
is?
Actually
you
know
organic
usage
in
the
ecosystem,
which
is
still
early
days,
so
I
think
that
the
answer
to
question
three
is
going
to
be
depressingly
small,
the
near
future,
but
it's
we
need
to
figure
it
out.
A
Akiva
brought
up
in
interesting
points.
Are
we
discussing
the
peg
today
and
we
did
kind
of
discuss
the
peg
briefly,
but
we
perhaps
we
didn't
come
to
any
kind
of
a
general
consensus
on
that
and
I
think
that
we've
seen
the
numbers
we've
seen
some
graphs
I
think
vishesh
dropped
out
right
before
our
ADA
yeah
right
before
I
asked
whether
he
had
anything
tasty
to
show
us
votes.
The
health
of
the
peg,
but
I
want
to
get
a
general
sense
from
every
before
we're
done
today.
D
A
A
All
right,
so,
let's
stick
with
the
chat
thing.
So
if
anybody
has
any
opinions
about
whether
we
need
to
initiate
another
governance
poll
or
whether
we
need
to
wait
to
see
how
the
market
shakes
out
now
would
be
a
good
time
to
do
some
informal
signaling,
we
can
continue
the
discussion
in
a
reddit
thread
and
hammered
out
over
the
course
of
the
weekend.
A
A
D
Yeah
sorry,
I'm
I'm
yeah,
just
in
one
get
the
noise
in
the
background,
but
yeah
I
was
just
curious
if
it
seems
like
if
collateral
requirements
were
to
be
increased,
that
might
be
another
mechanism
by
which
you
could
limit
dye
supply
and
wants
sort
of
an
indirect
mechanism
seems
like
it.
It
couldn't
help
with
the
peg
curious
effect.
That's
been
discussed
at
all.
A
C
At
least
at
least
from
my
perspective,
I
think
they
achieve
kind
of
a
similar
indirect
impact.
As
the
stability
fee,
I
mean
civility,
and
collateralization
ratio
can
kind
of
think
of
it,
as
in
vs.
I
mean
points
on
a
curve
where
you
can
kind
of
adjust
one
or
the
other
to
get
to
the
same
point
as
far
as
kind
of
user
experience
goes,
I
think
changing
collateral
requirements
would
be
much
more
unfriendly
for
having
like
the
same
ultimate
impact.
That's
how
I
see
it
is.
A
B
The
other
thing
you
ought
to
keep
in
mind
is
how
those
particular
parameters
are
calculated
in
the
first
place,
so
as
an
example,
debt
ceiling.
A
way
to
think
about
this
is
imagine
you
have
one
single
CD
PE.
What
is
the
maximum
amount
of
collectible
you'd
like
to
be
you'd
like
to
have
deposited,
given
that
you
wish
to
liquidate
it
over
a
given
time
period?
So
you
know
that's
how
you
answer
the
first
question
about
debt
ceiling,
then
the
the
second
one
about
liquidation
ratio.
B
That
is
very
much
related
to
obviously
online
volatility
of
the
asset
itself,
and
you
know
that
speaks
to
what
the
liquidation
ratio
should
be.
And
then,
of
course,
you
have
what's
left
over.
Is
the
stability
fee?
Now,
if
you
take
everything
else
out
of
the
equation,
then
the
stupid
CP
would
generally
be
very
small
because
it
should
be
able
to
cover
the
economic
cost
of.
B
Straight
it
has
a
fourth
component
called
the
liquidity
premium,
but
considering
we
don't
have
any
maturity
periods
on
these
on
the
use
of
dye
that
gets
taken
out
of
the
equation.
So
the
tricky
part
comes
in
here
where
the
stability
disability
fee
has
this
inflation
and
has
this
credit
component
and
if
you
take
it
away,
you're
left
with
the
operational
component
and
that
operational
component
serves
to
power
two
purposes
first,
to
cover
the
cost
of
running
an
operation,
but
also
at
the
same
time,
to
act
as
a
policy
instrument.
B
So
that's
why
you
want
to
use
the
stability
fees
specifically,
you
can,
if
you
wish,
not
change
the
others
and
create
a
policy
component
in
there
as
well.
But
then
what
you
do
is
you
create
permutations
and
that
creates
complexity.
So
it
kind
of
needs
to
first
come
from
the
principle
of
the
first
principle
basis
of
understanding
what
a
debt
ceiling
is
truly
supposed
to
achieve,
and
that
is
always
in
the
context
of
the
whole
system
is
robust.
B
Everything
is
moving
forward,
full
bull
and
it's
working
perfectly
as
opposed
to
where
we
are
now
where
we're
saying
well
things
of
bootstrapping.
So
as
an
example,
if
you
have
a
debt
ceiling
of
let's
say
you
know:
fifteen
million
dollars
on
token
ABC
and
you
know
the
whole
d5
space
is
robust,
everything's
moving
forward
appropriately.
What
you're,
basically
saying
is
that,
if
you've
hit
that
debt
ceiling,
you
need
to
go
find
more
collateral
types.
B
There
isn't
this
option
of
increasing
it
anymore,
because
your
risk
parameters
have
definitely
definitively
concluded
that
anything
above
that
is
a
risk
to
the
system
that
we
cannot
and
we
do
not
wish
to
bear
where
we
are
right
now
in
bootstrapping,
we
are
slowly
getting
to
these
true
debt
ceilings.
So
keeping
that
in
mind
there
is
this
option
to
think
well,
whilst
bootstrapping,
why
not
just
use
the
debt
ceiling
as
part
of
this
mechanism?
You
could
but
then
you're
working
against
the
general
ethos
of
why
it
was
there
in
the
first
place.
B
If
you
have
a
look
at
the
liquidation
ratio,
then
you
basically
actually
all
you're
doing
is
literally
transferring
the
risk
out
of
the
stability
fee
into
into
the
liquidation
ratio.
So
the
profile
as
a
whole
stays
the
same
to
two
Cyrus's
points,
but
if
you
start
attribute
it
to
different
risk
parameters,
you
just
create
a
little
bit
of
a
complexity
that
that's
very
difficult
to
manage.
B
So
that's
why
the
stability
fee
is
the
best
place
to
you
know
deal
with
this
particular
risk
and
to
also
deal
with
this
particular
dislocation,
because
it
aligns
appropriately
an
increase
in
stability
fee
both
affects
the
supply
and
demand
function.
A
decrease
has
the
opposite
effect,
both
in
the
supply
and
demand
function.
So
it
does
have
a
coherent
and
complete
functionality
that
is
appropriately
incentivized
just
by
using
a
stability
fee.
D
That's
great
thanks
and
just
one
other
comment
to
leave
and
don't
need
to
discuss
further,
but
I
think
gonna
be
great
on
the
call
to
hear
from
various
liquidity
providers
in
the
ecosystem.
You
know
and
I'm
assuming
they're
you
know,
Joe
is
probably
going
to
be
great
too.
So
you
see
if
we
can
engage
them
in
the
call
and
get
their
responses
since
they're
the
ones
that
are
likely
to
be
most
impacted
by
an
in
these
changes.
Yeah.
A
A
It's
it's
something
that
I
want
to
continue
and
that's
one
of
the
things
like
cyrus
touched
loosely
on
a
working
group.
That's
been
formed
and
some
of
the
major
members
of
that
are
liquidly
providers.
We're
still
wrestling
with
these
issues
around
transparency,
because
obviously
people
have
businesses
to
run,
and
they
also
have
competing
and
complementary
incentives
to
provide
accurate
data.
A
So
we
need
to
we're
still
feeling
feeling
out
those
complexities,
but
I
think
that
that's
gonna
be
a
core
component
of
these
calls
in
the
future
is
some
easy
way,
whether
it's
just
our
liquid
providers,
giving
us
a
thumbs
up,
emoji
or
a
thumbs-down
emoji,
or
whether
we
have
a
very
sophisticated
snarks
based
to
all
where
people
could
submit
their
numbers,
and
we
can
try
and
verify
it
that
way,
but
yeah
we
need
to.
We
need
to
understand
who
the
primary
stakeholders
are
in
this
ecosystem.
Yeah.
B
This
falls
into
that's
that
notion:
I,
don't
know
who
brought
it
up
about
dipole.
You
know
based
dipole.
Basically,
you
know
using
the
idea
of
LIBOR
in
in
a
decentralized
space.
That
is
the
essence
of
a
to
to
JC's
question.
We
are
literally
asking
all
liquidity
providers
to
come
in,
but
you
know
they
need
to
be
incentivized
appropriately
to
to
contribute
what
they
know
and
we're
trying
to
align.
B
That
incentive
by
you
know
simply
saying
keep
in
mind
that
you
know,
if
you,
you
know
state
clearly
what
your
position
is,
then
you
can
also
state
clearly
what
you'd
like
to
see
the
stability
fiat
as
a
contributing
member,
because
it's
not
going
to
be
it's
not
going
to
be
terribly
unclear
as
to
you
know
why
they
wants
it.
If
you
got
what
you
consider
to
be
high
inventory,
you
may
propose
a
very
large
stability
fee.
B
Then
it's
just
a
case
of
going
out
what
we
know
why
you
want
the
large
stability
fee,
but
at
the
same
time
we
need
to
just
gauge
how
gauge
the
level
that
you're
proposing
maybe
pull
that
back
a
little
bit.
You
know
maybe
control
it
from
everyone
else's
point
of
view.
That
is
essentially
the
the
premise
of
this
kind
of
liable
proposition
where
you
do
have
decentralized
contributions,
where
you
clearly
know
where
the
incentives
lie
in
those
contributions
and
submissions
all.
A
Right
we
I
thought
we
were
not
going
to
have
enough
questions
and
I
don't
have
to
kill
time.
But
of
course
we
have
too
many
questions
and
not
enough
time
fish.
You
posted
some
really
great
graphs
and
I'm,
sorry
that
we
didn't
get
a
chance
to
view
them
together,
but
I'm
going
to
make
sure
that
they're
called
out
specifically
in
the
discussion
thread
that
I'm
posting
right
after
this
call,
because
I
happily
have
a
window
in
my
schedule
to
do
so.
A
I'm
also
going
to
capture
the
rest
of
the
chat
messages
that
weren't
addressed
because
there's
some
good
things
in
there
Aviv
at
the
very
end.
You
you
pitched
this
idea
with
us
doing.
Another
poll
I'd
like
to
explore
that
a
bit
in
the
Reta
thread
as
well,
so
keep
an
eye
out
for
that,
because
it's
it's
very
important
to
me
that
people
in
this
call
specifically
are
engaged
with
the
maker
governance.
Some
Brit
subreddit,
because
that's
where
a
lot
of
the
science
is
getting
dropped
and
there's
some
really
interesting
discussions
there.
A
B
Yeah
yeah.
Well,
what
I
want
to
basically
just
say
again
is
everything
that
we've
considered
today
has
been.
You
know
from
the
the
point
of
view
of
community
members
contributions
right.
Basically
one
one
emphasizes
this
is
exactly
what
we
need
to.
You
know,
make
a
decentralized
platform
work
appropriately.
Also,
the
other
thing
to
keep
in
mind
is
that
we
are
facilitating
this
as
the
foundation
so
that
we
can
gather
this
information
in
order
to
help
with
this
momentum.
B
But
please
be
please
be
very
clear
on
this
notion
that
it
is
this
decentralized
aspect
that
is
the
most
important
point
behind
scientific
governance
and
we
are
contributing
to
this
phase
in
terms
of
Felicity,
as
in
terms
of
facilitating
this
as
quite
clearly
the
kind
of
a
superuser,
that's
also
a
non-voting
member.
This
is
very
critical
for
everyone
to
understand
so
that
we
can
gauge
clearly
who
is
involved
in
how
so
and
to
that
end.
I
was
looking
at
the
folks
on
this
call.
B
We
hit
a
roundabout
run
about
61,
to
63
and
of
those
63
I'd
say
about
80%
of
the
folks
were
non
foundational
members.
So
you
know
that
really
sort
of
bodes
well
that
we
are
pushing
this
forward
appropriately
into
a
property
centralized
platform,
and
thank
you
very
much
everyone
for
participating
again.
Please
keep
on
contributing
and
please
be
cognizant
of
the
the
the
governance
mechanism
and
the
voting
mechanism
and
contribute
where
you
can.
A
F
A
A
We
need
to
get
out
of
the
trap
of
being
crypto
correlated,
which
was
unknown
risk
and
that's
part
of
our
beta
stage,
but
we're
definitely
going
in
that
direction
and
I
might
be
stepping
on
Cyrus's
toes
because
I,
don't
like
I've,
said
numerous
times,
I'm
no
expert,
but
it's
it's
custody,
its
kose
AML
and
it's
recourse
that
are
we
in
your
considerations
when
we
talk
with
these
things
and
to
put
it
into
a
tight
little
package,
so
makers
now
can't
do
any
of
those
things.
That's
just
not
a
right.
A
A
Well,
no
I
think
it
would
be
an
option.
I
think
that
the
organization
is
moving
to
a
place
where
we're
going
to
be
able
to
support
things
like
that,
or
at
least
the
our
side
of
things
like
that.
But
the
question
is
still
outstanding
of
how
much
the
maker
doubt
can
alter
its
core
infrastructure
to
support
securities
and
how
mature
the
the
ecosystem
becomes
to
make
it.
So
we
don't
need
to
write
God's
Word,
we're
in
the
spectrum
things.
A
Land
I,
think
that
stos,
like
our
the
utility
token
of
2019,
we're
kind
of
hoping
that
this
comes
along
and
saves
crypto.
We
don't
know
when
it's
going
to
happen,
how
quickly
it's
gonna
happen
and
how
robust
is
going
to
be
when
it
gets
here,
but
their
gold
maker
does
to
be
prepared
for
it
as
soon
as
it
gets
on
the
radar.
Sorry
to
do
in
my
own
face
on
this
thing.
You
know.
A
C
A
Know
yeah
we're
actively
anxiously
fine
looking
for
partners
that
can
provide
those
things
too.
It's
really
the
common
example
is
always
real
estate.
Like
hey,
can
I,
you
know,
stick
my
condo
in
a
CDP.
The
response
to
that
is
probably,
if
you
find
an
enormous,
Lee,
complex
and
sophisticated
and
well-run
organization
that
will
handle
that
tokenization
yeah
we'd
love
to
see
that's
open,
but
we're
not
in
the
business
of
you.
A
F
C
Yeah
I
I
messaged,
you
back,
but
so
yeah
I
mean
there's.
There's
a
lot
involved.
Online
I
have
some
documentation
in
the
works
on
specifically
what
ok
cool
maker
risk
teams
looks
like
and
and
what
their
roles
are
on,
how
they
would
interact
with
the
community
with
foundation.
If
they
were
to
need
to
do
that,
or
you
know
compensation,
that's
I
mean
the
thing
is
a
lot
of.
It
is
still
kind
of
somewhat
of
work-in-progress
I
mean
when
I
say
we're.
I
mean
we
there
hasn't.
C
Some
decisions
have
been
finalized
yet
on
how
we
think
the
best
way
to
go
about
it,
but
I
mean
it
really
doesn't
even
matter
what
we
think
is
the
best
way.
I
mean
the
risk
team.
Ultimately,
you
know
either
very
unlikely
does
work
for
free
just
out
of
whatever
and
they
just
submit
risk
work
or
they,
you
know,
somehow
apply
with
the
MKR
holder.
You
know
somehow
need
to
create
some
sort
of
communication
meeting
where
they
can
interact
directly
with
some
amorphous
group
of
em
care
token
holders
and
go
through
their
yeah.
F
A
Think
that
we're
very
very
deeply
in
the
bootstrapping
phase
when
it
comes
to
external
risk
teams
and
so
I
think
that
part
of
the
process
of
figuring
out
how
that
process
works
will
involve
heavy
amounts
of
input
from
people
that
were
interested
in
risk
teams.
So
I
think,
though,
for
a
lot
of
stuff
we're
gonna
be
looking
to
people
like
you,
Baron
or
other
people
that
are
interested
in
eventually
spawning
off
their
own
teams
and
work
on
that
process
together.
Yeah.
F
Definitely
I
was
for
seeing
something
like
we
would
build
an
algorithmic
approach
to
assessing
risk
and
collateral
parameterization
for
those
risk
parameters.
That
can
be
something
that's
developed
like
on
chain
where
people
would
submit
something
that
we
can
pull
an
on
chain
price
history
for
look
at
the
volatility.
Look
at
things
like
the
best
friend
exchange
availability
and
basically
set
kind
of
like
a
not.
F
It
doesn't
necessarily
have
to
be
like
a
machine
learning
algorithm,
but
a
trained
algorithm
to
assess
what
those
risk
parameters
on
a
very
conservative
basis
would
need
to
start
out
as
and
then
improve
over
time,
so
wouldn't
need
as
much
manual
oversight
and
can
be
done
more
seamlessly
over
time.
I
think.