馃攳
I Interviewed a Registered Futures Algo Trader | Algorithmic-Futures - YouTube
Channel: Jacob Amaral
[0]
okay so thank you kyle for hopping on
[1]
today and to talk about automated
[3]
trading systems and building algorithmic
[5]
futures kind of more about yourself so
[8]
i'm here with kyle schultz who runs
[9]
algorithmic futures which provide
[11]
trading systems that allow you to auto
[13]
trade your strategies mostly with
[14]
futures in your inner trading account so
[16]
he runs that he has multiple systems and
[17]
he's going to talk about that uh later
[19]
in the in the video but kyle tell me
[22]
more about yourself kind of where you
[23]
came from how you got into automated
[24]
trading and what you do
[26]
yeah absolutely i kind of i'll dig into
[28]
my background i've got a unique kind of
[30]
pathway into algorithmic trading and
[33]
i've done a lot of different things
[34]
within finance investing to get here
[37]
um but by way of background i'm kyle
[39]
schultz i'm managing director of
[41]
algorithmic futures
[42]
like you mentioned we offer automated
[44]
trading systems um originally from
[47]
chicago so uh if you know anything about
[49]
chicago it's got it's kind of the heart
[51]
of prop trading it's got tons of prop
[53]
trading firms that
[54]
all were born out of the exchanges there
[56]
this the board of trade chicago to trade
[59]
and then the cme right so
[61]
you know growing up my i had my brothers
[64]
friends who
[65]
uh traded at different prop trading
[67]
firms and just seemed like the best job
[69]
in the world to be quite frank right
[70]
because
[71]
intellectually stimulating
[73]
um you're only working market hours
[75]
right so they could be golfing by 3 p.m
[78]
and you know at the end of the day they
[79]
made
[80]
did pretty well uh and so i kind of saw
[82]
that and gravitated towards it
[84]
and just started getting more
[85]
interesting technical analysis
[88]
and prop trading in general but kind of
[90]
figured out i had a lot to learn
[93]
um coming coming out of undergrad at
[95]
university of iowa it was 2008 and a lot
[97]
of those prop trading firms stopped
[99]
hiring traditional traders and really
[101]
just wanted quant traders and developers
[104]
to be specific
[106]
so i didn't have that type of background
[107]
so i found it difficult to get into a
[109]
prop training shop
[111]
but ended up in investment consulting so
[113]
we were allocating capital for different
[114]
pension funds endowments and foundations
[117]
which was really interesting and i was
[119]
doing manager research both for you know
[121]
traditional stock picking long equity
[123]
managers but also hedge funds
[126]
um and you know obviously hedge funds
[128]
were much more related to the prop
[129]
trading world so i took a deep interest
[131]
in that and it was a good place to cut
[133]
my teeth
[134]
um and then ultimately ended up at
[136]
jackson national which is a 200 billion
[138]
dollar insurance company um so i was
[140]
allocating capital with different hedge
[142]
funds there
[143]
but because we managed so much money we
[146]
really could only allocate capital to
[147]
the largest hedge funds the aqrs
[150]
the brevan howards the bridgewaters of
[152]
the world
[153]
and so i became fascinated with the cta
[155]
space
[157]
i ended up going back to school at
[158]
university of california los angeles for
[161]
my mba
[162]
finished the cfa program
[165]
and started to learn a little bit of r
[168]
from my data analytics classes and that
[170]
kind of sparked a renewed interest in
[172]
trading for me
[174]
i always had the mindset of a
[176]
quantitative trader i just didn't have
[177]
the skill set so once i picked up r i
[180]
got really into back testing different
[182]
types of strategies
[184]
and ultimately developed a cta and then
[187]
algorithmic futures
[189]
um to focus on trading um
[192]
you know i found it difficult to break
[194]
into the industry industry to be
[196]
completely honest because i didn't i
[198]
wasn't spinning out of a large hedge
[200]
fund or anything like that i didn't have
[202]
the pedigree that
[204]
a lot of top um ctas and
[206]
and traders have um so i really just
[209]
developed my own cta and started
[211]
fundraising
[213]
and developed different strategies in
[215]
the
[216]
futures markets and options on futures
[219]
and the cta space is very
[221]
similar to the systems trading space
[223]
i'll explain the differences uh later on
[226]
and when we get into algorithmic futures
[228]
uh but that was kind of my progression
[230]
towards ending up in in trading and
[233]
running a cta and then algorithmic
[235]
futures
[237]
super cool yeah you have quite a quite a
[239]
unique background because a lot of
[241]
people
[242]
that get into the space maybe they come
[243]
from a computer science background or
[246]
um some type of math degree and then
[248]
they maybe start at a hedge fund so it
[250]
sounds like you know you kind of
[252]
you having that business background and
[254]
then being able to apply that to the
[256]
insurance company you worked at and
[257]
managing capital with with um hedge
[259]
funds that's super super cool um one one
[262]
thing i just wanted to clarify if you
[263]
guys don't know what a cta is it stands
[265]
for commodities trade advisor just so
[267]
you guys know uh some people may not
[269]
know people might think it means click
[270]
to action or something like that or call
[272]
to action
[273]
so i just wanted to clarify that um
[275]
super cool kyle awesome um and and i
[278]
wanted to kind of harp more on
[281]
how did you get into like traded and
[283]
automated trading specifically like what
[285]
was the year were you trading in college
[287]
were you trading in high school when did
[288]
that start
[290]
yeah so as i mentioned my brother's
[292]
friends were prop trading i got into
[294]
kind of more traditional equity analysis
[297]
the tough part is like when i went to
[298]
school in undergrad they didn't have
[301]
financial engineering degrees at least
[302]
for undergrad they had kind of master's
[304]
programs that started to develop after
[305]
that but you know they weren't really
[307]
teaching prop trading and trading in
[310]
general so i gravitated a little bit
[312]
more towards traditional equity research
[315]
asset allocation
[317]
portfolio optimization and just more of
[319]
the
[320]
traditional academic route so like but
[323]
all of my experience um in in trading
[326]
and developing trading systems has been
[327]
more grass roots right so like you
[330]
mentioned i don't have a quad background
[332]
per se i don't
[333]
i didn't study uh computer programming
[336]
or do a masters in financial engineering
[338]
i had a little bit of data analytics
[339]
that i picked up during business school
[341]
and i kind of ran with that and really
[343]
learned through
[344]
iteration on stack overflow right so
[347]
yeah you know i start to develop stuff
[349]
and then go to stack overflow and just
[351]
kind of learned it by myself
[352]
um
[353]
and was was kind of utilizing r both r
[356]
and python for my back testing engine
[359]
until i discovered tradestation
[361]
and tradestation easy language um
[364]
so that kind of gives you a background
[366]
about how i how i kind of got into quant
[368]
trading even though i don't have a
[369]
traditional
[371]
computer development background or
[373]
really it's a super quantitative
[374]
background i understood the trading very
[377]
well because i was interviewing hedge
[378]
funds and allocating the hedge funds
[380]
but i kind of picked up the development
[383]
piece uh by myself
[385]
cool that's so cool yeah i'm a complete
[387]
opposite i came from a computer science
[390]
background i didn't know much about
[391]
trading and so i kind of had to
[394]
find that
[395]
that happy medium of you know yes i know
[398]
how to code stuff but you also know how
[399]
to think like a trader uh
[401]
diversification make sure your
[403]
strategies are correlated and stuff like
[404]
that make sure the strategies are built
[406]
correctly and then we'll go more into
[408]
that so um you already you already
[409]
mentioned it my next question was what
[410]
broker and software you use so you did
[412]
say use tradestation um and obviously
[414]
they have the easy language um
[416]
uh programming language if you will
[419]
and i guess did that kind of force you
[421]
to go into futures like what made you
[422]
start going more into futures than than
[424]
equities
[426]
yeah so there's a couple reasons the the
[428]
original strategy that kind of sparked
[431]
my interest in in back testing and
[432]
starting my cta was
[434]
i was focused on trading iron condors i
[436]
don't know if you're familiar with that
[438]
but it's an option
[439]
strategy where you're selling call
[441]
spread on a put spread with the same
[442]
maturity uh works best in sideways
[445]
markets
[446]
um you know in down markets and up
[449]
markets usually you've got to rebalance
[451]
the position and you might for example
[453]
in a down market you might rebalance and
[455]
take a little bit of a loss but when you
[457]
reset the position
[459]
it's typically at higher implied
[460]
volatility so you get more premium in
[462]
the future right so there's
[464]
it's a strategy where it works well in
[466]
sideways markets but can also do well in
[469]
in bear markets like we're seeing this
[471]
year uh especially if it's not
[474]
it's not a very violent down market it's
[476]
more just kind of a downward trending
[478]
market like we've been seeing or like
[479]
2001 and 2002.
[482]
so um started back testing
[485]
you know option based strategies and
[487]
then when i wanted to register for a cta
[490]
ctas are specifically focused on the
[493]
futures market
[494]
uh so i really started to hone in on on
[497]
those markets and there's a lot of
[499]
benefits right you essentially get
[501]
free leverage through margin trading
[504]
so in the stock market you might have to
[507]
pay a cost to borrow to your brokerage
[509]
in the futures market everything is
[511]
traded on margin and so you're
[512]
inherently have leveraged positions but
[514]
there's no cost to that leverage
[517]
um so there's a lot of
[518]
there's a lot of capital efficiency
[520]
within the futures markets
[522]
and then you also have exposure to a
[524]
variety of different markets whether
[525]
it's commodities interest rates
[527]
currencies um or equity indices which
[530]
can be very uncorrelated
[533]
cool super cool uh yeah i like i like i
[536]
i like that you mentioned the kind of
[538]
sort of the free leverage if you will
[539]
when it comes to futures obviously you
[541]
have to have the capital to be able to
[542]
enter that position right the cme sets
[544]
the rules
[545]
but um
[546]
that is one of the biggest benefits and
[548]
personally i love futures too because
[549]
it's so easy to diversify and to like
[552]
build a system for indexes and then
[553]
build an oil system or a gold system um
[557]
obviously with equities there is etfs
[558]
for sure but usually the liquidity is a
[561]
bit lower right like if you look at oil
[563]
and and natural gas etfs if there is any
[565]
uh the volume is a bit lower and just
[568]
really a lot easier with futures so
[570]
that's that's one of the main benefits
[571]
that i talk about as well on my channel
[573]
um okay that's awesome so my next
[576]
question is um
[578]
i know it's a little bit of a doozy but
[579]
and there's many more best practices but
[582]
when when more specifically when
[584]
building automated trading systems um
[587]
we'll pick three for now but what are
[589]
like the three best practices for
[591]
building automated trading systems
[592]
because anyone can code one anyone can
[594]
go into easy language or personally i
[596]
use ninjatrader and go into ninja script
[598]
and you know buy on monday sell on
[599]
friday or buy on an estimate crossover
[601]
what are kind of like the three best
[602]
practices that you would say help you
[605]
find success in automated trading
[608]
so the biggest issue in quant trading is
[611]
overfitting right um
[614]
with quantitative trading you can create
[616]
whatever strategy you want based on
[618]
different signals and parameters
[620]
but ultimately you really have to if you
[623]
don't focus on robustness you're
[626]
trading out a sample is not going to be
[628]
as good as the back test and so a lot of
[629]
people fool themselves into thinking
[631]
they have great trading strategies when
[633]
they've just over fit the data to a
[635]
specific time period
[637]
so so a couple different ways to avoid
[639]
that
[640]
is one
[641]
don't use 10 20 parameters right focus
[645]
on a handful of parameters whether
[646]
that's two or three i find that that's
[649]
kind of the right number
[651]
because the more parameters you add the
[653]
more complexity you have in your model
[655]
uh and then the more you're specifying
[657]
each of those parameters to the to the
[659]
market
[661]
so i would stick to only a couple
[663]
different optimizing a couple different
[664]
parameters
[666]
and then two uh what i would do is i
[668]
would use walk forward analysis
[671]
and specifically within tradestation
[673]
there's cluster walk forward analysis so
[675]
in walk forward analysis you're using
[678]
different time periods of and how to
[679]
sample and in sample period typically um
[683]
in sample you can use 70 80 percent then
[685]
out of sample use 20 thirty percent what
[688]
a cluster walk forward analysis does it
[690]
actually varies the in sample and out of
[693]
sample period over the walk forward so
[696]
if you have a ten ten year time period
[698]
that you're uh running a back test on
[701]
you might cut that into
[703]
two two year periods and you have an in
[705]
sample out of sample and you're walking
[707]
it forward every two years to understand
[710]
if those parameters that you're using
[713]
are robust
[714]
and i you know i can go into a lot more
[717]
detail right tradestation has specific
[719]
um parameters to determine the
[722]
robustness and efficiency which you can
[724]
tweak
[725]
uh but ultimately you're trying to make
[727]
sure that those parameters you know if
[729]
if they change significantly from period
[731]
to period you probably don't have a very
[734]
robust uh
[735]
back test and going forward you're not
[738]
going to experience the same type of
[739]
performance that you did in the past
[743]
so those are those are kind of two of
[744]
the main things um
[746]
and then also i think the third thing is
[749]
you want to look and inspect you want to
[751]
back test over
[753]
and this is controversial but you want
[755]
to back test over long periods of time
[756]
and the reason i do that is because i
[759]
use the back test because i want to
[760]
inspect each volatility event
[763]
or each period that the strategy did not
[765]
perform well and really actually
[767]
logically understand what was happening
[769]
look at the trades
[770]
look what was going on in the market
[771]
that at the time was it you know
[773]
inflationary period like we're
[775]
experiencing now where interest rates
[776]
increasing decreasing how much
[778]
quantitative easing was there um because
[781]
if you're just back testing over a
[783]
specific regime take for example uh 2010
[787]
to 2018 right
[789]
period of massive qe interest rates were
[792]
zero there's no inflation
[794]
um when you have a regime change
[796]
your strategies aren't going to hold up
[798]
probably
[799]
and i think it's what we're seeing with
[801]
a lot of quant strategies this year uh
[804]
because the market regime uh this year
[806]
is much different than it has been in
[808]
the past so i do like to look at very
[810]
long term back test i'm not necessarily
[812]
optimizing to the longest time frame but
[815]
i'm i'm researching and analyzing um the
[818]
data over that time period
[821]
perfect yeah i would totally agree with
[822]
what you said i i definitely follow some
[824]
of the same steps as you i think
[826]
specifically with walk forward one of
[828]
the rules that i've recently developed
[830]
that's helped me kind of cut out some
[831]
bad strategies is
[833]
sometimes when you test your initial
[834]
idea and then you run your walk forward
[837]
um and then say you have a test period
[839]
of you know maybe a year or two years
[841]
and then your optimized period is like
[842]
two years for example
[843]
if the test period accounts for more
[846]
than 50 of the total net profits of the
[848]
walk forward test usually that's a red
[850]
flag because that means
[852]
in that year just your strategy was did
[854]
very well but it could not perform
[856]
better in the other test periods right
[858]
um and vice versa if your parameters if
[860]
you slightly optimize them and your walk
[863]
forward changes a lot right like maybe
[865]
the net profit goes down by 50
[867]
uh or even up and there's so much
[868]
volatility that's usually a red flag of
[870]
like hey your strategy's not too robust
[872]
so i like that a lot and then also
[874]
talking about the volatility periods i
[875]
think that's big too because
[877]
um if you understand how your strategy
[880]
performed in a certain event
[882]
um for example say an 08 or covid or the
[885]
red years of 2015 and 2018
[888]
uh i think it helps you
[890]
also help build future strategies to
[892]
complement it right so for example maybe
[895]
um you have a seasonality strategy that
[897]
trades oil and it does well in these
[898]
months but not the well in the other
[900]
months well you can build a strategy to
[902]
complement those other months uh trading
[903]
a different asset on a different time
[905]
frame sort of thing so um i think those
[907]
best practices are are very sound and
[909]
something that you know i follow as well
[911]
so that's really cool um
[913]
i want one thing one thing i also want
[915]
to say is like you bring up a good point
[917]
like each volatility event is very has
[920]
very different characteristics right so
[922]
what you saw happen in covet incredibly
[925]
violent and quick down market
[927]
and then recovering pretty significantly
[929]
and quick
[931]
is much different from the time period
[932]
we saw in 2008 uh and and also from if
[936]
you go back to 2001 2002
[939]
we had just kind of a deflating of the
[941]
tech bubble that occurred over a
[943]
two-year period right and so the
[946]
the depth and and and the speed of of
[949]
volatility events is much different so
[951]
when you understand how the strategy
[952]
strategies performed during each of
[954]
those volatility events you have a
[955]
better perspective for going forward
[959]
yeah i agree kobe was very violent
[961]
downside and then upside where 08 took a
[964]
longer time it took multiple months to
[966]
recover i think over a full year and
[968]
then 2000 and 2001 were a lot longer too
[971]
so
[972]
um
[972]
that's something to take into account so
[974]
very interesting stuff i love it um
[976]
i would definitely agree with those best
[978]
practices so uh my next question is
[981]
um you may be able to talk more about
[984]
this but um a lot of my my audience and
[986]
my viewers you know they have a
[987]
nine-to-five job and then they you know
[989]
they build their own trading systems on
[990]
the side and they run them um you know
[992]
either on their local computers or say
[994]
on like a vps or something like that
[996]
what would you say are some tips and
[998]
tricks to
[999]
you know i know automated trading
[1001]
systems are automated but is there any
[1003]
tips that you can recommend for like
[1005]
running automated systems if you have a
[1006]
job already
[1007]
and things to kind of watch out for
[1009]
maybe if you're new
[1011]
yeah absolutely so um
[1014]
and just to kind of elaborate on on what
[1016]
you were mentioning is like
[1017]
after business school i also worked in
[1020]
private equity and work in m and
[1022]
investment banking right so
[1024]
um i've done a lot of different things
[1026]
uh within finance and investing um
[1029]
but i've always had a passion for
[1031]
training that i've started i've
[1032]
developed my training systems really on
[1034]
this side and i did the cta for about
[1037]
two years full time but ultimately
[1040]
um
[1041]
had trouble with the fundraising side of
[1042]
the business and that's why the systems
[1044]
business is great because i can use
[1046]
digital marketing advertising to bring
[1049]
in clients and really just kind of go
[1051]
out there on the internet set up shop
[1053]
and and everything's automated right so
[1056]
um
[1057]
it's very scalable and can be run um
[1060]
can be run really with not a lot of
[1062]
hours per per week or per month focused
[1065]
on just more marketing
[1067]
now that business i have a broker
[1069]
partner called the fox group and what
[1072]
they do is they handle all the
[1073]
onboarding
[1075]
and all of the administrative aspects of
[1077]
running the system so they monitor the
[1079]
systems all day like you said even if
[1081]
you have automated trading systems
[1082]
things go wrong
[1084]
whether you're using ninjatrader or
[1085]
tradestation or whatever it is you have
[1087]
to you have to keep an eye on it um
[1090]
but there's there's there's ways to stay
[1093]
in tune with what's going on with your
[1095]
trades well i get notifications to my
[1097]
phone and honestly i know my strategy so
[1099]
well that based on the time of day or
[1101]
what contracts being traded i know what
[1103]
strategy it is right
[1104]
um so
[1106]
you know technology like that helps but
[1108]
i do have a broker that's full-time
[1110]
monitoring all the systems and all my
[1112]
customers accounts
[1113]
um and then they also help with rolling
[1116]
futures contracts
[1118]
um and doing trade allocations and all
[1120]
that so so algorithmic futures is a
[1123]
really nice uh systematic trading
[1125]
business because i've got a broker um
[1128]
that's kind of handling all the
[1130]
um administrative aspects and then i can
[1132]
focus on developing strategies and
[1134]
marketing them online so a couple of
[1136]
tricks you know is is is definitely to
[1141]
use a vpn like you said if you if you
[1143]
have like a remote desktop like whether
[1145]
it's aws or another system it just helps
[1148]
reduce interruptions to your trading
[1149]
systems and you can kind of log in
[1152]
whenever you need to to check on things
[1155]
and then you know also the frequency of
[1157]
your trades it's important you know if
[1159]
you're
[1160]
incredibly active you probably have to
[1162]
monitor the trades a little bit more to
[1164]
make sure everything's trading correctly
[1165]
make sure there's no issues but if
[1166]
you're doing for example long-term tread
[1169]
following where your hold periods are
[1170]
are weeks days or weeks
[1173]
it's probably easier just to automate it
[1176]
you have less buys and sells and there's
[1178]
there's less um
[1180]
less surveillance and monitoring of the
[1182]
systems required
[1184]
cool i like that yeah i definitely
[1185]
recommend vps i use aws myself and
[1188]
it just saves me all the hassle like if
[1190]
my power goes there my internet goes out
[1192]
there's actually a condo building being
[1194]
built near my office and sometimes the
[1196]
power goes out like not recently but
[1198]
like about a month ago because they were
[1199]
probably hooking stuff up and if my
[1200]
system was running on my local computer
[1202]
uh it'd be turned off right so but it
[1204]
was running on aws so that was not an
[1206]
issue so yeah definitely um and that's
[1208]
really cool that you work with that
[1208]
broker partner i think we'll talk about
[1210]
more um about that at the end um and and
[1213]
one of the benefits that if people go um
[1215]
work with you in algorithmic futures um
[1217]
they can see that so that's really cool
[1219]
um okay cool uh my next question is and
[1222]
the second last one um
[1224]
what are your
[1226]
so
[1227]
there's intraday and and long-term swing
[1229]
trade training strategies right so
[1232]
uh there's pros and cons to both um i
[1234]
find that some intraday systems can have
[1237]
more noise and more randomness however
[1240]
you can usually have
[1242]
less risk higher sharpe ratios less
[1243]
stress running intraday systems and you
[1246]
know exiting before the the settlement
[1248]
session um but swing trading systems um
[1251]
can be also very stressful especially if
[1253]
you're always in a position so what are
[1255]
your thoughts on you know developing
[1257]
systems for intraday versus swing do you
[1259]
want a mixture of both um how do you
[1261]
handle the risk to reward on those two
[1263]
types of kind of strategies
[1264]
[Music]
[1266]
yeah so i like to and algorithmic
[1268]
futures portfolios are really
[1269]
diversified between uh what you would
[1271]
call swing which i think of as like days
[1274]
to weeks holding period and then
[1276]
intraday strategies
[1278]
um the market the market microstructure
[1281]
and the market patterns are can change
[1285]
for intraday periods and also by weeks
[1288]
and months right so you might have
[1290]
periods where intraday strategies
[1292]
perform incredibly well and your swing
[1294]
strategies are not performing well it's
[1296]
just another layer of diversification in
[1297]
my opinion
[1299]
the benefit you do get from intraday
[1301]
trades is you tend to have
[1303]
a larger sample of trades in your back
[1306]
test which is helpful um when you're
[1308]
thinking about robustness um
[1311]
and then swing trades you know
[1314]
there might be fewer trades over a month
[1317]
or years or over the back test period
[1319]
you're looking at so you've got to be a
[1321]
little bit more skeptical with it and
[1323]
maybe just tweak some of your
[1325]
so what some of your um thresholds for
[1328]
what you consider a robust strategy in
[1330]
my opinion so i like i like the intraday
[1332]
strategies because there's a larger
[1333]
sample size um but ultimately combining
[1336]
those strategies in a portfolio provides
[1337]
another layer of diversification so you
[1340]
can think of diversification from like
[1342]
okay what markets are you trading
[1344]
commodities
[1345]
uh equity indices treasuries currencies
[1349]
um types of strategies momentum mean
[1352]
reversion and time period of strategy
[1354]
intraday or swing right so the more the
[1358]
more diversification in my opinion that
[1360]
you're able to put into the portfolio uh
[1363]
the better the better it's going to
[1364]
perform over longer periods um and the
[1367]
less periods of drawdowns that you're
[1368]
going to experience
[1370]
cool okay so it overall it's just better
[1372]
than both if you can and just the more
[1375]
time you spend developing eventually you
[1376]
get to a point where you'll have some
[1377]
intraday some swing systems um i what i
[1380]
like to add is too is looking at
[1382]
specific times of the day
[1384]
there's actually lots of research papers
[1387]
talking about just reading one the other
[1388]
day of like if um
[1390]
you know correlation with overnight
[1392]
returns versus you know the first 20
[1394]
minutes or the last 30 minutes of the
[1396]
session um now the the research papers
[1398]
that i read were on uh equities
[1400]
specifically but i think when it comes
[1402]
to intraday systems um maybe do some
[1405]
like probability and like statistics on
[1406]
okay you know what happens in the first
[1408]
30 minutes of the day if the overnight
[1410]
returns are positive or negative how are
[1412]
they correlated and then building a
[1413]
trading system around that
[1415]
and then obviously you know does it
[1416]
perform in the walk forward but um
[1418]
i recently i've been developing more
[1421]
intraday systems because i i sort of
[1423]
started with a lot of you know longer
[1424]
swing training systems like you said
[1426]
days and weeks and i would say those are
[1429]
sort of easier to build because um
[1433]
there's less noise
[1434]
but the problem is sometimes when you're
[1437]
actually running them live
[1438]
and they're not going your way because
[1439]
they take days or weeks to exit
[1442]
it is sometimes painful and you want
[1443]
those intraday systems to maybe make
[1445]
some some quick returns to kind of
[1447]
offset that um but anyways that's more
[1449]
of the the psychological aspect of
[1451]
running these systems so
[1453]
yeah no that makes a lot of sense right
[1454]
you have to understand
[1456]
you have to understand you know your
[1458]
risk threshold and how you're going to
[1460]
react to these strategies and if if
[1462]
you're not going to follow the systems
[1464]
then you've got too much risk in the
[1466]
strategies or it's not the strategy for
[1468]
you right so
[1469]
um it's one thing to run a back test and
[1472]
see a drawdown and a drawdown duration
[1474]
on paper
[1475]
yeah yeah it's a different thing to to
[1477]
live through that right so um those are
[1480]
just i think
[1481]
you know if you're starting out trading
[1483]
uh doing paper trading and then getting
[1485]
into live trading will help um help you
[1488]
kind of with your emotions and sticking
[1490]
to the systems right so you can develop
[1492]
great systems but you also have to stick
[1494]
to the strategies um and be able to
[1496]
write out the drawdowns
[1498]
that's right yeah and i know a lot of
[1499]
the back tests i do i just look at that
[1501]
the net profit parameter i'm like sweet
[1503]
it made you know 100k over 10 years that
[1505]
means i'm going to make it and it's just
[1507]
not the reality sometimes you have to
[1508]
let the systems play out and uh
[1511]
one of my early mistakes too was kind of
[1513]
overriding my systems by closing the
[1515]
position manually um just because i
[1517]
couldn't handle it and i think that just
[1519]
says that like i wasn't able to handle
[1521]
that strategy the drawdown and you have
[1522]
to be able to realize that
[1524]
and i usually recommend
[1526]
when you're building systems like
[1528]
having them all in some type of an excel
[1530]
sheet or some type of like organizing
[1532]
tool to say hey these are all my systems
[1534]
here are the max drawdowns maybe the
[1536]
average monthly profits that sort of
[1538]
stuff and that way you can kind of
[1540]
um prepare yourself and yeah paper
[1542]
trading too i would say that any
[1543]
especially any new system you build you
[1545]
probably want to paper trade it for a
[1546]
period of time um just you can get
[1548]
accustomed to it so um that's that's
[1550]
really good stuff kyle thank you
[1551]
appreciate it um my last question before
[1553]
we talk about algorithmic futures um
[1556]
what are some
[1558]
realistic goals that you know automated
[1560]
traders can set like what do you
[1562]
personally set in terms of are you
[1563]
looking at a yearly return basis what's
[1565]
your return to drawdown like what kind
[1566]
of goals can people set that are
[1568]
realistic and that help them kind of
[1571]
achieve and then kind of motivate them
[1573]
to get past the hard times and the good
[1574]
times
[1576]
yeah i think i think that's a good
[1578]
question because um because
[1581]
when you get interested in trading or if
[1583]
you're just starting out and trading
[1585]
sometimes you're looking at back tests
[1587]
or you're looking at some of the top
[1588]
hedge funds or other traders that you
[1590]
know
[1591]
and it's very difficult to untangle
[1593]
skill and luck in trading unless it's
[1595]
over a very long period of time right
[1598]
so you could have had a trader that did
[1600]
incredibly well in 2020 and they're just
[1602]
getting crushed this year right so
[1605]
it's very difficult and it takes a long
[1607]
period of time to really see
[1610]
if you're a skillful trader or you just
[1612]
you know got lucky in a specific market
[1614]
right
[1616]
um so i i think
[1619]
and
[1619]
sorry going back to the question i i
[1621]
think there's there's a lot of different
[1623]
things you have to think about but using
[1625]
if you look at return to max drawdown
[1627]
usually you want to target like a great
[1629]
strategy would just be a two cal mark
[1632]
that's a kmr ratio is reached annualized
[1634]
return to max drawdown um so if you're
[1637]
able to develop a strategy that does uh
[1640]
30 annual return with a 15 max drawdown
[1644]
i think that's a great threshold and
[1646]
even if it's uh you know in between a
[1649]
one to two kel mart that's realistic
[1651]
right um and when you combine them in a
[1653]
portfolio
[1654]
you might get some benefits of
[1655]
diversification to reduce that max
[1657]
drawdown so that's one of the benefits
[1659]
of having multiple different types of
[1660]
strategies in a portfolio is like you
[1663]
might have um a couple different
[1666]
one one-to-one kelmar ratio strategies
[1669]
but they're uncorrelated when you
[1671]
combine them in a portfolio that reduces
[1673]
your max drawdown increases the kalmar
[1675]
ratio of the total portfolio to two
[1677]
um i think that's one thing to focus on
[1680]
in the institutional world or within um
[1684]
within the h1 world a lot of people use
[1686]
sharp ratio but there's there's
[1688]
definitely some drawbacks to sharpe
[1690]
ratio because
[1691]
that's annualized return to standard
[1693]
deviation and it really doesn't
[1695]
differentiate between upside deviation
[1697]
and downside deviation
[1699]
um so i think understanding which risk
[1702]
return metrics uh work best for you or
[1705]
how you think is important but i like uh
[1707]
return to max drawdown and if you're
[1709]
trading with the leverage max drawdown
[1711]
is what you care about right because
[1712]
you're you're your whole goal is to have
[1715]
a capital efficient portfolio
[1718]
and optimize the return so that you know
[1720]
for me like i try to create portfolios
[1723]
where
[1724]
you're not going to get a margin call
[1725]
now there's always risk of margin call
[1728]
but from my back testing the way i set
[1730]
up my portfolios
[1732]
you're able to use the least amount of
[1734]
capital
[1735]
to not get a margin call and then
[1737]
increase returns right so it's kind of
[1739]
an optimization game with futures and
[1741]
capital
[1742]
um but yeah if you can develop one to
[1744]
two kelmar
[1746]
ratio uh type strategies you're you're
[1749]
looking in the right place and
[1750]
developing the right types of strategies
[1753]
awesome i would totally agree i think
[1755]
um
[1756]
that's sort of the best realistic goal
[1758]
you could also go simpler and say hey
[1760]
i'm gonna develop you know five to ten
[1762]
systems this year and these markets and
[1764]
those are very achievable goals and then
[1766]
um the the end goal is to you know have
[1769]
a a profitable year right because that's
[1771]
the whole goal to make a positive net
[1773]
return obviously you can set that goal
[1774]
it's really hard to predict that you
[1776]
never know what's going to happen but um
[1778]
as long as you're you know doing these
[1779]
best steps that we talked about you know
[1781]
walk forward analysis diversification um
[1784]
the cluster walk forward analysis that
[1786]
you talked about too and building
[1786]
systems right um i think you're you're
[1789]
headed in you're the ship is in the
[1790]
right direction right so um i love that
[1793]
cool awesome um okay well appreciate it
[1795]
kyle so i wanted to uh before we wrap up
[1797]
beer i want to talk more about uh
[1799]
algorithmic futures which is your kind
[1801]
of product for automated systems um
[1804]
i'll be linking in the description below
[1805]
as well in the comments where they can
[1807]
find you but talk about more of the kind
[1809]
of the systems you have and and some of
[1811]
the the products you offer
[1813]
yeah absolutely so i manage a website
[1816]
called algorithmic futures.com there's a
[1819]
dash in there where i offer uh different
[1822]
trading portfolios that i've created to
[1824]
clients and like i said we use a broker
[1826]
partner with all the onboarding and to
[1829]
handle all the administrative aspects of
[1831]
setting up the account rolling futures
[1834]
contracts and whatnot so it's a hundred
[1836]
percent automated hands-off trading and
[1839]
essentially everything that i've
[1840]
developed which is really five different
[1843]
portfolios
[1844]
focused on
[1846]
intraday momentum intraday mean
[1848]
reversion
[1849]
and then swing trading momentum and mean
[1852]
reversion and trend following um it
[1855]
combines all those strategies in
[1856]
different portfolios ranging from
[1858]
uh account sizes as small as 15k up to
[1861]
65k right so the larger the account size
[1865]
the more strategies you can add and
[1867]
different types of strategies so they
[1868]
tend to perform better the larger the
[1870]
account size you have um but you know
[1873]
it's a way for
[1875]
for customers who are interested in
[1878]
trading maybe they're a discretionary
[1880]
trader and they've just had a really
[1882]
tough time and they're starting to
[1883]
understand that hey look i like trading
[1886]
um and i want a good uh return on
[1888]
capital but
[1889]
you know i can't i can't i can't focus
[1892]
on the market 24 7 and maybe they
[1894]
utilize algorithmic futures as a portion
[1896]
of their portfolio
[1898]
or you have um different types of people
[1900]
that trade long only equities and bonds
[1902]
which have not done very well in bear
[1904]
markets right
[1906]
they want to they want to diversify into
[1908]
types of strategies
[1910]
that can outperform in a bear market
[1912]
and you can go to the website and look
[1914]
at some of our live performance and back
[1916]
tested performance
[1917]
to see how strategies have behaved
[1919]
during different volatility cycles and
[1921]
events
[1922]
most of this most of the systems are up
[1924]
and doing very well this year and the
[1926]
systems focused on trend falling within
[1928]
commodities have done incredibly well
[1929]
this year um so yeah i'd urge people to
[1932]
go to the website poke around and
[1934]
understand how how we develop our
[1936]
systems and see if they're interesting
[1938]
uh for you or even just to
[1940]
break off a portion of your portfolio
[1943]
and allocate to those types of
[1944]
strategies
[1946]
perfect
[1947]
awesome yeah so once again that that
[1948]
link will be in my description and
[1950]
comments below that's algorithmic
[1952]
futures.com um definitely give that a a
[1955]
look out i think um you have some really
[1956]
cool systems there and you definitely
[1958]
added some new ones too because i
[1959]
remember checking out your website um a
[1961]
couple months ago so it's cool you're
[1962]
adding new systems uh in the future so
[1964]
that that's that's really good to see um
[1966]
that's all i had kyle thank you man i
[1968]
appreciate your time i appreciate it
[1970]
yeah we had some good chats and lots of
[1972]
gems and i think a lot of value in this
[1974]
video so um thank you so much and um we
[1977]
hope to see you soon
[1979]
yeah thanks for having me it was a great
[1980]
discussion and uh if anyone has
[1982]
questions feel free to reach out to me
[1984]
i'm glad to talk to any other traders
[1985]
and um and thanks for thanks for the
[1988]
interview
Most Recent Videos:
You can go back to the homepage right here: Homepage