馃攳
Billions: Solving Taylor's liquidity challenge with Quantitative Trading - YouTube
Channel: unknown
[0]
we're getting into hot water about
[1]
market liquid i'm gonna dump five
[2]
million shares of blood horn steel at
[4]
40.
[4]
problem is trades by appointment matthew
[7]
needs to sell two hundred million
[9]
dollars of this blue horn steel stock
[11]
he can't execute the order because there
[12]
isn't enough liquidity in the market
[14]
no broker will take all 5 million and
[16]
guarantee my price there's been a hot
[18]
debate in the comments on how to solve
[19]
this
[20]
we need a final answer i'm getting help
[22]
from the leading experts in outdoor
[23]
trading
[24]
if you wonder why we're talking about
[25]
algorithms to solve a liquidity problem
[28]
watch till the end but first what is the
[30]
problem you can see on bluebird the bid
[32]
offer spread is two cents
[33]
but that only works for small amounts
[36]
liquidity represented here by the order
[37]
book looks like this chart
[39]
the bid offer is only valid for a
[40]
specific size for larger trades the
[43]
spread will widen
[44]
and it's moving all the time matthew
[46]
wants to sell 5 million shares above 40.
[48]
he can't simply sell it in the market he
[51]
can't ask a broker
[52]
nobody will guarantee the price taylor
[54]
offers a solution
[55]
what if you create a zero cost collar
[57]
for more details about the strategy
[58]
there's a video about it
[60]
right use options to protect the sale
[63]
buy forty dollar puts in blue dorn then
[65]
sell forty five dollar calls to offset
[67]
the purchase
[68]
i said that according to bloomberg the
[69]
option can't be executed
[71]
the volumes are too thin the liquidity
[73]
problem can't be solved with options
[75]
celestial outcomes said i'm reading
[77]
bloomberg wrong it could be set up
[78]
differently
[79]
maybe but we have to give some credit to
[81]
matthew if that was the case then he
[83]
wouldn't have a liquidity issue in the
[84]
first place
[85]
the second point is that they could
[86]
trade with primes directly it's called
[88]
over the counter or ltc
[90]
but the issue is the same as for the
[91]
spot trade who will price for that huge
[93]
amount in options
[94]
and placing the orders otc also means
[96]
i'm telling the market what my position
[98]
is
[99]
they will talk then i got hit by rob
[101]
kimball's comment
[102]
we could try to conceal the trade over
[103]
time it's a subtle strategy using the
[106]
caller for a portion of the amount while
[107]
selling as much as possible in the spot
[109]
market
[110]
and conceal the option trade over
[112]
different expires but this is a risky
[114]
business
[114]
what he means by 14 times adb can be
[116]
spotted here
[117]
the average number of shares traded
[119]
daily is 350k
[121]
what about the initial solution from
[122]
mafia so i've got five brokers
[125]
selling a million each and they don't
[127]
know about each other can we pass on the
[128]
problem to someone else
[129]
get five people to do it at precisely
[131]
the same time are you sure about the way
[133]
you're unloading that
[134]
the issue is that brokers will check
[135]
prices before dealing that size kind of
[137]
shady and not very efficient
[138]
it's likely they will figure it out if
[140]
they know there's more they will move
[142]
their bid
[142]
and worse they will quickly get there is
[144]
a huge amount to be sold
[146]
the whole market will know it's a
[147]
visibility issue none of this works
[150]
to find the solution i interviewed a top
[152]
expert in the field
[153]
and a pioneer in electronic trading joe
[155]
wald he's built and sold companies
[157]
dealing with that
[158]
and he's got a precise solution here is
[160]
how he explained the problem the
[162]
marketplace
[163]
today is highly fragmented and trades in
[166]
very small size so
[167]
average trade size on an exchange is
[169]
probably 100 shares or less and the
[171]
average order size of an institution is
[174]
10 000 shares or so he talks about large
[176]
institutions but the problem is exactly
[178]
the same as the one facing acts capital
[181]
and this is how to solve it electronic
[182]
trading and algorithmic trading is
[184]
really the methodology
[185]
in which institutional clients leverage
[188]
their
[188]
access to executing a trade when you've
[190]
got a fragmented market where you've got
[193]
multiple places where you can execute
[195]
and you also have high frequency trading
[197]
trying to get an edge based on what
[200]
they're seeing in the marketplace
[202]
it becomes an incredible challenge for
[204]
an institutional client to be able to
[206]
execute their orders without signaling
[209]
or creating
[210]
a lot of market impact an institutional
[212]
client that's looking to to trade
[214]
electronically
[215]
to execute a large block is leveraging a
[218]
strategy
[219]
depending on what they're looking to do
[220]
potentially a liquidity sourcing
[222]
strategy that's going to work the order
[225]
based on a quantitative kind of arrival
[227]
price kind of
[229]
strategy it's also going to look to see
[231]
if there's any
[232]
large blocks of liquidity that take
[235]
place in some of the
[236]
the dark pools or some of the venues
[238]
that are at midpoint
[239]
two important concepts here first
[242]
fragmentation means that you need to
[243]
look for liquidity in many places
[245]
you can't do that manually it needs to
[247]
be electronic and algorithmic
[249]
the second dark pools are venues where
[252]
large investors can trade without being
[253]
visible
[254]
unlike what is happening on exchange so
[256]
what would happen if axe capital was a
[258]
client of joe's and was passing the
[260]
order to his team
[261]
now we've got this order we understand
[263]
the client's objective
[264]
they've chosen a strategy of hours that
[267]
they're going
[268]
to use and it's our job to make sure
[270]
that the strategy performs optimally and
[272]
that
[272]
we're picking the right venues and the
[274]
right order types
[275]
and watching what's happening in the
[277]
marketplace really closely
[279]
to make sure that we're executing as
[282]
best we can
[282]
and ultimately delivering a great
[284]
outcome for the client so there you go
[286]
mafi
[286]
problem solved are you happy now or are
[289]
you scared you will be replaced
[290]
just like in tommy knockers first it
[292]
makes things better for you
[293]
then it destroys you this is how
[295]
computers take over the world thanks for
[297]
watching
[297]
next time i promise it will be the real
[299]
me speaking before you go please don't
[301]
forget to subscribe and like the video
Most Recent Videos:
You can go back to the homepage right here: Homepage