Up next in 10
Join Leo in this captivating podcast interview as he dives deep into the world of trading with the exceptional Kevin Davey, a three-time championship futures trader, and algo trading maestro.
Uncover the fascinating story of how Kevin transitioned from a factory manager to developing his own trading strategy factory.
As a prolific author on algorithmic trading and a mentor to a thriving community of algo traders, Kevin shares valuable insights on the significance of risk management, diversification, and crafting personalized trading strategies to conquer the financial markets.
Don't miss this opportunity to learn from a true mastermind in the realm of trading!
✉️ Learn the secrets of legendary traders and investors:
https://bit.ly/4aCna7i
🔗 Follow along:
https://analyzingalpha.com/episode-004-kevin-davey
⏱️ Video Chapters
0:00 - Introduction
1:18 - How Kevin got started
5:50 - Kevin’s first strategy
7:43 - Backtesting in Fortran
Show More Show Less View Video Transcript
0:00
you're about to enjoy Alpha cast episode 4 where I have the absolute pleasure of
0:05
interviewing Champion Trader Kevin Davey he's a champion because he's not one not just one or two but three first and
0:12
second place finishes with multiple triple digit returns in a real money's future trading Championship he's an
0:19
algorithmic Trader with unique way of pumping out strategies using his strategy Factory approach if you're a
0:25
Trader and especially if you're an algorithm Trader you're going to love this one everyone Kevin Davey Hey Kevin
0:32
pleasure to uh to be here with you today I'm super excited uh as far as I'm concerned I think you might actually be
0:38
the most prolific algo trading writer on all of Amazon is that is that correct
0:45
um it might be I don't know but I I guess you know you're kind of prolific when you start to see other authors on
0:53
Amazon copy your book covers and try to mimic what you're doing and I've seen quite a few of that because I'm like boy
0:59
that looks like my cover and and they actually seem to sell pretty well so yeah you could be right no that's
1:05
fantastic that's fantastic well obviously uh for those of you who don't know uh Kevin Davey fantastic
1:11
algorithmic Trader we're going to get into his background how he got started and his advice for new and experienced
1:17
Traders so with that being said Kevin how did in the world did you get in them financial markets
1:24
um it really started right after college uh and probably a lot of people have
1:29
been down a similar Road I got something through the U.S mail at
1:34
the time now today it would probably be like hey you get an email or you see a YouTube video but it was uh basically an
1:41
advertisement for this campus trading system to trade sugar
1:46
and um I was hooked maybe I saw it and I was like wow this makes sense even though it
1:53
was you know just all Sizzle and there was really no stake there
1:58
that kind of started me and and just started me down this road of trading and
2:04
losing money and trying a different approach and losing money and then doing
2:09
something else losing money and then finally kind of getting to uh what they
2:16
call algo trading now which was systematic trading or mechanical trading back then
2:21
and which is just rule-based trading you know things like hey if it's a 10 bar
2:28
High breakout you buy and if it's a 10 bar low break out you sell short
2:34
and you put that in the computer and let it tell you how it did
2:39
and that kind of just got me down the path and once I started doing that I started getting uh a lot better at it
2:46
here we are perfect now did you ever get into discretionary training
2:52
yeah uh many times where I thought hey I'll just follow this price action Guru
3:00
who you know he does uh I remember one in particular did end of day analysis and boy he'd hit it all the time and it
3:08
was like wow this is like the key and then I went into his live trading
3:14
room for like a week and he was like well look at this it may go up from here
3:19
it may go down and you're like well wait it doesn't sound as great as you sound at the end
3:25
of the day and I realized that a lot of discretionary trading
3:31
um there's probably people who can do it really well but I'm not one of them and
3:37
it seemed like I'd buy right at the peak I'd sell right at the low point and uh
3:45
you know and then your emotions get involved so oh I'm gonna trade double next time to get it back and you start
3:50
doing a lot of goofy things so uh eventually I gave it up I can't say that
3:56
I'll never try it again because every once in a while I'll probably try something but for the most part
4:03
I just go pure algo trading now yeah no no that that's great um and sort of the I I like the maybe
4:10
not the concrete or the the fake um confidence that you get from a back
4:16
test right it feels to me at least hey I've done some rigorous work on this and I know that discretionary Traders they
4:21
do deep Dives and back tests too but I've always felt like I guess certain people are drawn to the the number side
4:28
of things and that was me whenever I was doing the discretionary stuff I sort of ran into the same issues as as you
4:34
effectively yeah yeah that's a good point and it's
4:40
funny because you know every time I'm talking to a great uh you know uh algo
4:45
Trader I can uh age their experience level by whether they say algorithmic
4:50
trading systematic trading or mechanical trading yeah so there's a lot of there's a few Market Wizards that you know I
4:56
talked to about this and they're like they talk about the you know mechanical training but anyways speaking of you
5:02
know mechanical training you know what you know when you first got into uh you know the algo trading uh did it you know
5:10
how did you get started did you you find you know what your current platform and just start working did you find you know
5:16
someone else that was doing and said you know what maybe this is for me because it's discretionary stuff isn't working out
5:21
well actually uh this kind of tells you how old I am but I started to do a
5:28
triple moving average crossover system and I did it
5:34
with pencil and paper and getting the prices out of uh I was living in
5:39
California at the time so I think it was either the LA Times or the Orange County register and so I'd get the prices and
5:45
write them down Cal had a hand calculator and calculated all by hand and I think I read about that uh in some
5:53
book or somewhere that hey this triple moving average system hey it works great and so
6:00
that was the first thing I actually did I didn't actually back test it because I
6:05
didn't have a computer to back test it with but I said hey I'm gonna start doing it
6:11
because you know look at look at this chart look at these examples looks like it works right I start out with the five
6:17
thousand dollar account and I still remember I was trading uh I think it was live hogs at the time now they're called
6:22
lean Hogs and I was taking every signal religiously and probably like within a
6:30
week or so I lost about fifteen hundred dollars out of my five thousand
6:36
I was like oh man and then I had the brilliant idea that if what I wasn't doing was working the exact opposite
6:44
must work right so then I started fading all those signals so when there
6:51
was a like an uptrend type thing I would go short and that started to work for
6:58
maybe a few days maybe a week I don't quite remember but then it hit a big
7:03
Trend in Hogs and of course I was on the wrong side and there goes another fifteen hundred dollars or so so you
7:10
know within a month or a couple weeks I forget I was out 50 60 of my money and you know
7:18
then I had to take a break from Trading again but that was how it had all started so I wasn't even back testing in
7:25
the beginning and um after a little while then I realized oh
7:31
I should probably be back testing and I was kind of a do-it-yourselfer
7:37
back at that point and actually programmed some things in uh actually
7:42
the language I was using was Fortran which you know that was like an old engineering language but it it did the
7:50
job and I was able to create a decent back tester of course I didn't know how
7:56
to back test I just know knew I should be back testing and you know time proved
8:04
that you got to know how to back test not only if you can do it
8:09
so it took a while took a while for sure yeah no no that's fantastic I haven't heard or used Fortran since my old
8:16
computer science day so it's it's been a while that's great okay so then you uh
8:22
so you built this uh the system and you you're saying now obviously it makes sense to know how to back test
8:29
um and I guess obviously as we all you know grow on the maturity curve of of
8:35
back testing expertise for lack of better terms you know what you know what should you have been doing back then
8:41
that you know that you should be doing now um a couple things one I had no idea of
8:49
the costs involved like with slippage so for example uh even on that the mini s p
8:57
you can pretty much assume that you're going to get a tick slippage uh
9:03
every time you go into a trade at least that's what I assume sometimes you don't get it and you get the get a good price
9:10
but a lot of times what your back test would say is hey I got in at this price
9:15
but really you got in that price plus a tick so you're already down
9:20
with the mini s p like 1250 per contract that was a big thing because uh what
9:28
I've found in the years after that is most ideas whether they're breakout
9:35
moving average whatever if you don't include trading costs in the calculations most things work out on
9:42
balance you know they're they make some money a moving average system is a great
9:48
example but as soon as you include trading costs then it falls apart and you know and a
9:56
lot of people say well hey that I'll do the reverse but you know trading costs work against you no matter which way
10:02
you're going and most times that's a killer so that's one thing another thing I really didn't
10:09
understand uh at least as far as programming and making sure it was right was doing uh limit orders so I was like
10:18
oh I don't want slippage so I'll use limit orders well what a lot of trading
10:23
software does and what I was doing at the time was hey as soon as my price was hit I'd assume that limit order was
10:31
filled and then you know the market would take off I'd be in it it was all good but the reality is most
10:39
times with limit orders your price has to be uh breached before
10:45
you get in so let's just say you wanted to buy it a hundred or better well the
10:50
price has got to go down to 99.9 to really guarantee you a fill it could hit
10:56
100 over and over and over and you might never get filled and then the market takes off guess what you're not in that
11:03
trade so just knowing some of those intricacies of the market and uh
11:09
were things when I was back testing at first I just didn't do and
11:15
I'd create systems that would seemingly take advantage of
11:20
and then what ended up happening is it it kind of bit me in the long run yeah no no that's all those are all good
11:26
points I mean you know when you start back testing the last thing you know and you know that you're trying to develop a
11:32
strategy the last thing as a new back test you're thinking about is frictions and data quality and all these other
11:38
things that you just assume oh yeah these are these are fine but we know from experience that it's that's not
11:43
necessarily the case right absolutely okay so
11:49
um Now Kevin I know that you've had a few run-ins with an investing championship and you've done pretty well
11:56
um can you give me uh some insight into what those investment championships were and and how you I guess masterminded you
12:02
got I think top top seed if I recall correctly right yeah I actually finished
12:07
uh in second place two years and in between I finished in first place all
12:14
those years were over a hundred percent return and uh that those
12:21
performances were what really convinced me that hey I might be able to do this full time and so to back up a little bit
12:30
I was slowly getting better at my trading I thought I was doing decently and and but you know you never really
12:38
sure if you're doing that well and so I saw this contest out there it's called
12:44
the World Cup of Futures Trading it was the same one that Larry Williams legendary Trader Larry Williams turned
12:50
10 000 into I think 1.1 million over the course of a year so it was real money it
12:57
was a long-term concert uh contest did he do that during a championship or
13:02
pardon did you do that one thousand I remember that run but did he do that through the actual championships yeah he
13:09
did it I thought it was in uh in the 80s at some point and then years later his
13:15
daughter who uh Michelle Williams who's actually the uh Academy award-winning
13:20
actress she actually did it I think using him as a proxy or his help and she
13:26
won it too so that was kind of interesting but the thing the one interesting thing that Larry wrote in
13:32
one of his books about that he he took ten thousand and ended up the year at 1.1 million but he mentions that he was
13:40
also at 2 million at one point so he actually lost about a million uh
13:47
you know he still did phenomenally well but you can get an idea of how much risk he
13:53
was taking on in that contest by doing that but uh you know so I knew about that contest
14:00
and I said well that's one way for me to gauge myself against other people so I
14:07
kind of took a year and I fake did it meaning I didn't actually enter it but I acted
14:14
like as if I did and I would have won that year and I was like oh man I should
14:20
have done it but you know how many people go through that and they you know they have this great performance until
14:26
it becomes real but that convinced me I said okay I can
14:31
do it and so what I decided to do was I said well I'm not going to set a goal for winning because
14:38
some years the people who'd win had a thousand percent other years it was a hundred percent
14:44
all over I said well I just want to finish in the top three
14:49
and so at the time this is before I entered I actually went to the people
14:55
running the contest and say hey do you have the top three finishers for the past five years or whatever
15:02
and they gave me the data and I came to the conclusion that if I hit a hundred
15:08
percent return in a year that I should finish in third place
15:14
or so or Ohio so I said that's great that's my goal so
15:20
I designed a system that took on a lot of risk but
15:25
it could end up with over a hundred percent return and then that's what I
15:30
ended up trading so uh I think looking back on it the the neat thing wasn't oh
15:37
hey he got over 100 return it was more hey I have this goal of a certain return
15:46
now I'm going to design some Algos to meet that and then I'll put it into
15:52
action let's see how it does because there's unfortunately there's so much noise in the markets that you don't know
16:00
if you're gonna actually do that you could get a run of bad luck and
16:06
you know you get caught in a drawdown and Things Fall Apart so
16:12
yeah you can't really have a goal and
16:17
expect always to meet it you just put yourself in the best position you can to meet it and then the Market's going to
16:24
do what it does and for me in those cases I was actually able to reach that goal three consecutive years so that was
16:32
the neat part about it no that's fantastic are you able to like dive into uh maybe not the exact details
16:39
unless it's an old strategy about how you actually reverse engineered hey this is a target I want to get to here's how
16:47
I think I can meet that I think that's fascinating well yeah so it
16:52
started with I said okay I just want to be a trend follower and I want to do a
16:58
few different markets to kind of be a little Diversified uh I was starting out with I think it was a fifteen thousand
17:05
dollar account so I couldn't be in too many markets at the same time but
17:11
that was the start and I picked some markets that I had lower margin requirements so I could be in maybe
17:19
three or four positions at a time and uh what I ended up doing was
17:25
creating a system and then running a lot of simulations and uh basically saying
17:31
okay well if this back test holds I have a pretty good chance of getting
17:37
over a hundred percent return yeah there's times where you're going to have a huge draw down but uh you should come
17:45
out on top if you don't get blown out but that's part of it is you know I took
17:50
the risk on it I said well there's a chance that my account is going to get blown out especially when you're going
17:57
for 100 returns you know you can't expect oh well I'm only going to have five percent max drawdown or ten percent
18:05
I mean people think that but that's not the reality the reality is you're taking on a lot of risk and I don't even recall
18:12
I think I wrote about it in my first book but uh I know I had a pretty big drawdown uh during that contest and what
18:21
ended up in the year I won what saved me was there was one it was one or two
18:27
trades and the one I remember in particular was copper where I just wrote a trend in Copper and without that I
18:35
definitely wouldn't have won but you know there's obviously I mentioned Randomness and luck and that does play a
18:43
role in some of those things which once I was done trading contests I kind
18:48
of convinced myself okay well I can kind of do this I tried to eliminate luck and
18:54
Randomness as much as possible but even today I still try to do it it's it's not
18:59
that easy to eliminate no no you can't force your will on the market although we all wish we could right yeah
19:06
definitely definitely wish we could yeah okay so in regards to uh you know your
19:13
strategy formations uh it sounds like to me when you're looking at a strategy you
19:19
know you obviously have a basic premise like a breakout strategy then do you optimize parameters over that or
19:25
um just trying to get an understanding of how whenever you're developing a new system you know what are you really
19:31
thinking about are you looking at the data first are you Theory driven empirical driven just trying to get an
19:36
understanding of if you're going to go set out to do a new system how would you how would you build that
19:42
okay well um over the years this is definitely my
19:48
Approach has definitely changed especially just with the computing power that's out there
19:54
now I tend to be a lot less theoretical and as far as putting tons
20:02
of thought into the idea of okay well I'm gonna have a breakout and when it pulls back then I'm gonna do this and
20:09
then I'm gonna have profit targets here and um that's the way I probably used to
20:15
think more but what I found is Simplicity usually is a better way to go
20:22
rather than try to have a system that does everything in you know a lot of
20:27
people like to have well I'll do this in a bear Market this isn't a bull market
20:33
um I just try to have some simple rules that work reasonably well over a long
20:40
period of time so that includes bull markets bear markets flat markets but it's not
20:46
necessarily optimized for each one of those what you want is you want it to
20:52
perform well in total but and you don't want it to fall apart
20:58
during any one particular time but I go I look at systems now where I can accept
21:04
flat periods for a long time that doesn't bother me as much as it used to
21:09
but back to the strategy development what I'll do is I'll create a system
21:17
with with a couple rules so let's just take a breakout system just as an example so you have your breakout length
21:24
uh and maybe I put a stop loss I have a stop loss in there that's two
21:31
different parameters that I could optimize uh and usually I like to stick
21:36
to under five let's say I'd like to use a lot of times two to four parameters
21:43
and that's all I'll optimized and I do a walk forward test which is a a way to
21:49
get out of sample data I tried to do is much testing as you need but you also
21:55
want to do as little as possible and what I mean by that is I don't run a walk forward optimization
22:02
over all my data look at the data and then say Oh I don't really like this
22:07
drawdown or I don't like this part the equity curve so I'm going to go and add a filter
22:13
I used to do that and it doesn't work it'll make a better back
22:19
test but that's not the goal you know the goal isn't to create something that
22:25
looks great historically it's to create something that looks decent historically
22:31
but that continues to work about the same in the future and that's where
22:38
most people get messed up in back testing because when you think about a back test
22:45
your output is that back tested Equity curve and so it's natural for people to
22:51
say well I want that to look as good as possible and
22:56
that's usually the exact wrong thing to think because you end up
23:02
adding filters so adding parameters over optimizing curve fitting doing all these
23:08
bad things and um it usually takes people a while before
23:15
they really get that you know I can tell people and probably people listening to this oh yeah okay don't over optimize
23:22
but then they'll go off and create back tests and still do it and it's not until you actually put real money to it to a
23:30
system that's over optimized and you watch it just fall in real time that you really understand it like oh maybe I did
23:37
something wrong here um but that's kind of in a nutshell what
23:43
my strategy development's about so I'll get ideas from anywhere whether it's a book magazine
23:51
internet YouTube video uh I'll just take it and code it up if
23:57
it's not already coded and then test it and then see and that's that's the best way so I
24:05
don't rely on anybody else to say well this works because there's a lot of people out there that claim their
24:10
special method works except when you test it and usually then it doesn't work yeah I found that to be true too and I
24:18
think some of it obviously I did some back testing on Candlestick patterns and you know for my YouTube channel I'll
24:24
back test whatever people can ask hey back test this or that and it's interesting that obviously
24:31
um the the information that's past generation to generation about what works is typically wrong unfortunately
24:39
and that might be just because the different markets you know but test everything would be my recommendation to
24:47
yeah it's that's the only way you feel confident in it in the long run
24:54
um that's why it's so hard to test other people's systems or trade them
25:01
because you don't know what all went into it but if you tested it yourself then you can have confidence and that
25:08
helps you to trade that live yeah yeah no that's fantastic and it's um you know I've had you're my fourth
25:15
guest uh and you're the third systematican all of them had said you know over time uh they've taken stuff
25:22
away the simpler is the better so so my next question to you is how many different strategies are you test or
25:29
trading it at once because usually that's the the recommended approach is to have a lot of simple strategies that
25:35
work together for diverse uh yeah diversification to me is the the biggest
25:42
thing to uh the Holy Grail uh I mean it's it's really good just to give you
25:47
an example like today gold is up a ton well I'm short gold whoops but
25:54
I have another algo that's long silver and so you know they kind of balance
26:00
each other out and that's what diversification helps helps you avoid some of those drawdowns and then over
26:06
time if you have 10 systems that over time make money
26:14
they'll still make money over time but now they kind of balance each other out with the drawdowns so that works great
26:21
and right now I'm trading live about 30 or so different strategies in in all the
26:28
major future Market sectors you know energies Metals interest rates and so on
26:36
and um that's out of a a group of about 175
26:41
strategies that I have that I could trade so I have a process to whittle that down
26:48
every month to say okay well I'm going to trade these uh the next month and
26:55
seems to work pretty well uh you know obviously there's times where I happen to pick the one that starts doing bad
27:02
and you know leave all the good ones on the sidelines but uh I have done at
27:09
times where I've traded as many as like I think 70 or 80 systems at one point
27:15
and um I mean that can be done you just run into more issues and that's the one
27:23
thing about automated trading that a lot of people don't realize they think hey I just set it up let it run automated and
27:30
come back in a week and you know look at how my account's bigger and that would be nice but uh you know there's just
27:37
little Internet outages I said one uh about a half hour before we're talking
27:43
where it was like a split second thing but sometimes that messes up some of
27:48
your automation so you got to be able to check it and um there's a lot of things that can go
27:54
wrong so you got to keep on top of it but again that kind of goes back to having
28:00
multiple strategies is you're not relying on just one and if that goes bad
28:07
you're sunk you know trading 30 hey some
28:12
will lose money in any month I get that but what you hope is that the ones that
28:17
make money make enough to carry the rest of them and a lot of times that does
28:22
happen you know that that's that's fantastic and I like the point of uh you mentioning simple strategies that you
28:29
can understand I went down the road of uh trying to use AI neural networks to
28:34
trade and I eventually got something working and then what I found out is that like I had no idea what it was
28:40
doing so I couldn't like actually have the confidence to keep it going right so that's just one of those things where
28:47
like you know people think that oh yeah the systematic Traders you can just turn it on and walk away and whatnot but we
28:54
still I don't know maybe you don't after these years but I still have um you know that maybe not as much of the pressures
29:00
as a discretionary Trader but you know we're still battling ourselves in the market right yeah you know it's funny
29:07
because I can always tell who's a a real automated algo Trader
29:14
if they say up there's no emotion go to algo trading there's no emotions and you
29:20
don't have to worry about that and as soon as they say that then I know either they're just Sim trading or you know
29:27
they're not trading at all soon or otherwise because as long as you have money on the line
29:33
you're going to have emotions and yeah it's with Algos it's not should I take
29:39
this trade necessarily but it'll be things like oh should I turn
29:46
this strategy off because it's lots of money the best six weeks and I kind of
29:52
lost confidence in it so those become emotional decisions sometimes and
29:59
you've got to realize that as long as you're dealing with money there's going to be emotions it's just going to be
30:04
different than what you're used to if you were like a discretionary Trader right but
30:10
um yeah that's uh kind of an understated part of it and I think it all goes back
30:17
at least my knowledge back to probably about 10 years ago maybe a little longer
30:24
when uh 4X was really popular and a lot
30:29
of people were doing it with metatrader they talked about these expert advisors and these robots oh just let the robots
30:36
trade and I mean I'd see ads for that everywhere and they always said oh no
30:42
emotion you just set it up and let it go and I can't imagine how many people lost
30:48
money doing that because it's just that's a tough way to go yeah yeah it's
30:54
funny because I feel like I've made every mistake in the book I mean my first algorithm ended up becoming so
31:00
complex and so over optimized by the end of it I didn't even remember what it was doing right I'm like oh this academic
31:05
paper looks good I can just do this portfolio after this that would be perfect and then like yeah at one point
31:11
I'm just like I have no idea what I'm doing uh so I had to throw it out but yeah it's it's funny I think I'm still a
31:17
few years behind you but I'm naturally gravitating to the to the fewer strategies I mean more strategies fewer
31:24
complexities let's put it that way yeah one thing that I'm not doing though is I I'm pretty uh predominantly a stock
31:31
Trader um so you do you do futures um can you tell me a little bit more about why you selected Futures over
31:38
stocks and quite frankly for me it wasn't a real big um thoughtful decision I started from a
31:45
valuation perspective and realized valuation and investing I like the trading more so that's why I've just
31:51
always been in the stock market was in a purposeful decision how about you and your futures decision
31:57
um so there's a couple things with Futures one is just the amount of Leverage uh you know obviously with
32:03
stocks you can buy stocks on margin and so you can get some leverage that way or you could trade options on stocks and
32:10
get a lot more leverage but Futures just the nature of them hey uh they're
32:17
highly leveraged products which obviously can work for you because you
32:22
can start out with smaller accounts but it can work against you so you know I I mentioned today uh gold is up uh it last
32:32
time I checked it was around Thirty thirty five dollars a contract which would be
32:37
thirty five hundred dollars and to put on a gold contract right now is
32:43
somewhere around eight thousand dollars in margin I believe so right then you could say wow I made
32:50
30 40 on my money today which you know most stocks you can't do that so while
32:58
it's a double-edged sword The Leverage is always appealed to me the uh tax treatment for people in the
33:04
U.S for futures is so much simpler than stocks um one you literally get a statement
33:11
from your broker at the end of the year with a bottom number and that's the only number you need to transfer to your tax
33:17
forms you don't have to do well the buy and sell and you know match everything up so that's huge
33:24
another uh part of that is there's uh what they call 60 40 tax treatment
33:30
where uh instead of you know with stocks if you hold it less than a year it's
33:36
considered short-term capital gains and you got to hold it more in the air for long term well Futures even if you held
33:43
it for one second uh either 60 or 40 percent I always get
33:49
them backwards but uh a certain percentage is considered long-term capital gains I mean for a one second
33:57
trade it's crazy so you can really do well tax wise uh with Futures Trading
34:06
and then uh you know the big thing I think Beyond those is you get to
34:13
be in different Market sectors so you know I mentioned some of them before but of course I didn't mention eggs and
34:21
softs and currencies and you can have systems and all those and you can also be in the stock market because there's
34:27
stock index futures and you can have a nice portfolio that's
34:33
in all these other all these things could you do it with stocks probably with some ETFs you could now because
34:40
there's a lot of more commodity based ETFs but uh I found that at the in the time I
34:49
started that really wasn't a possibility uh you know ETFs weren't even a thing way back then so
34:57
the the ability to diversify is a lot nicer in Futures and one other thing
35:02
I'll throw out which I've seen a lot of people who try to algo trade with stocks is they tend to bring their biases with
35:10
them meaning I'm going to create a I'll go for Amazon
35:16
oh okay I wonder what kind of algo it's going to be it's probably going to be pretty long most of the time for the
35:22
past 10 years because frankly that's what worked and so you know they already
35:28
know in the back of their head that what has happened and then they build an
35:33
algo that just kind of confirms it so it's it's not really uh you know a
35:39
non-biased kind of good algo in that sense uh you know what
35:45
would be good is you just took a random stock and you had no idea what it is and
35:51
you build an algo for that um but you know most people can't really do that but
35:56
you can do that with futures you know if I asked you hey what's gold done the past five years or if you ask
36:03
me that I would be like well it's probably gone up it's probably gone down I don't know
36:10
uh because those Trends in Futures tend to be
36:16
up there's a lot more up and down motion and that's actually good for eliminating
36:22
biases When You're Building Systems well that's great yeah I think uh some
36:27
good things to think about especially from that from the tax treatment a lot of people don't know that and obviously
36:33
the the tax man is sometimes not that friendly right you know so any Edge you
36:39
can get there is fantastic now uh so I guess at this point you know you've you've won the Championships
36:46
um establish a name for yourself you know I'm not sure exactly when you wrote the books but you've also started um I
36:52
think a community too you've you know I've seen some strategy Factory and in a community and a lot of people working
36:58
with you yeah so so what happened was uh after I had won the trading contest I was just
37:05
kind of off trading on my own and people would find me and so that sort of
37:10
started this little side business but um what happened uh it was more than 10
37:17
years ago now but uh a book publisher Wiley publishing there a big publisher
37:22
in financial books they approached me and said hey do you want to write a book and I thought oh okay well that's kind
37:29
of cool I always wanted to write a book and then you know I felt that way until I wrote it and then I'm like oh my gosh
37:35
I can't believe I just spent all this time writing a book to get like three or
37:40
four dollars per book royalties you know it's it's book writing is uh tough uh you know I
37:47
guess unless you're uh fiction writer that is a bestseller most people don't
37:53
make money selling bucks but I wrote the book and then which really was great because it helped me
38:00
uh just put all my thoughts down and it helped me even refine what I was doing
38:06
because if you you have to explain it in a book you've really got to understand it and so that helped out a ton
38:13
and uh as part of that I said okay well I spent all this time writing this book
38:18
and not what and as it turned out when the book came out people started coming to
38:25
me and say okay well I see your book but I really want to learn one-on-one with you or you know do you have a class do
38:32
you have a course and that started what became the what I call the strategy
38:37
Factory course and it's a course on building trading systems and a proper
38:44
way to do it and the reason I call it a factory is uh one
38:49
for some of my career I was in charge of quality assurance for a factory and so I
38:55
kind of obviously was on the factory floor a lot but that's what strategy building is you
39:02
take your idea that's your raw material and then you bring it in your factory and you run tests on it that would be
39:09
like if you had metal and you were running machines to hone the metal and shape the metal
39:16
and then at the very end you come out with the finished product which is either a strategy or it's junk and so
39:25
it's a perfect analogy for how you want to do it especially if you want to build multiple ones because what's a factory
39:30
do what's it good at it's good at building and building and building
39:35
and so that name just felt so appropriate to what I was doing because I was this building strategy after
39:43
strategy after strategy and that's what I teach in my course and so now after
39:50
doing it for a number of years we have a community of people we have a club where
39:55
people submit strategies and then I vow uh I evaluate them in real time
40:01
and it's like a six month task so there's no way to cheat and then if they're good people share
40:07
them which is really cool because I've never seen anybody else share actual strategies that hey this has worked the
40:14
last six months uh you know guaranteed because an impartial judge was watching
40:20
it and um so we just have this this whole group of
40:26
us who just build strategies and share strategies and so it's worked out pretty
40:32
well it's definitely made me a much better Trader because people will come to me and say Hey how
40:39
do you do this can you explain this and so then I add a new video to the course
40:44
and um you know I actually have one tomorrow I did uh over the over it was last year
40:53
I did over 50 million back tests uh all kinds of markets all kinds of
40:59
strategies and I analyzed all the data to kind of say well what things in back
41:06
tests help show that a strategy is going to make money in real time and so now
41:14
after all this analysis I'm sharing this and a lot of it's what I've already taught people but there's some new stuff
41:21
in there that will help people build strategies even faster so it's kind of kind of neat it's uh it's one it's me
41:29
teaching but it's also me learning and getting a lot better as a Trader as I go along so it helps me out too so it it's
41:37
uh pretty fun yeah no that's fantastic and I'm getting sort of the same from
41:43
some of my Trader friends they'll they'll say hey can you back test this for me just you know on Twitter and
41:48
whatnot but it's interesting because you notice when you have to do this stuff and then explain it you have at least
41:54
for me there's another filter okay I have to really be able to explain this in a manner that's not you know
41:59
non-techy and and how it actually works so that's great right and then in regards to your your
42:05
factory because I know a little bit about your trading systems but I think it's important to uh for you know
42:12
traders to understand is you know what process you know does each of the step of the factory for a strategy take
42:18
because I know that you have a pretty rigorous process to identify you're doing Walk forward analysis and a lot of
42:23
Traders should be doing some of these things but may not know that they exist or just might not know how to do it
42:29
okay so my process is seven or eight steps it starts out with goals you know
42:38
you've got to know what you want so that on the back end you can know if you've met your goal or hey should I go
42:45
back to the drawing board and test more so you've got to know that then you've got to have your trading idea and
42:51
obviously you want a whole bunch of ideas uh you want the stockpile hey I've got
42:57
all these things I've got to test so those are kind of preliminary steps just to get things going
43:03
then the the next thing I do is uh one of the things I do since most strategies
43:09
fail it's pretty much a given no matter what you're going to test no matter what
43:15
great idea you think you have chances are it's not going to work uh it's literally it's 99 times out of
43:24
100 that's true or as many as I've done some studies where it's one time out of
43:31
ten thousand it'll work you know something will work so most of the things you're going to
43:37
test are garbage so what you want to do is you want to limit the amount of time you spend on those so one of the steps I
43:44
do is I do a quick test with just a couple years of data and usually doing
43:51
that quick test you can not say oh this strategy is good but you can easily say
43:58
this strategy is probably not good uh you know there's no guarantee maybe you
44:03
picked a two-year period where the strategy just was bad and everywhere else it made tons of money but
44:10
usually not so that's an important step because it just helps you filter out
44:15
some of the real bad ones and then once you think hey I sort of
44:20
have something that works that's when you do your real testing and the first one I do is a walk forward optimization
44:27
and for those of you who don't know what walk forward optimization is compared to
44:32
a traditional regular optimization
44:38
don't have time to explain it here but definitely read up on it because walk
44:43
forward is probably second best in my book compared to just trading live you know
44:50
nothing beats trading live for proving something works but walk forward is probably the second best uh thing out
44:58
there and you know then you have things like Auto sample testing and probably the worst
45:05
uh next to not testing at all is to do an optimized back test
45:11
um that's usually don't do that that's usually pretty bad
45:17
so you run this walk forward back test and obviously there's certain ways to do
45:22
it there's certain rules to do it and you do it over all your historical data
45:28
and then you know you have an equity curve a this is what it looks like then I do uh a simulation a random simulation
45:36
I call it it's called Monte Carlo simulation that takes your trades on your Equity curve scrambles them all up
45:43
and gives you some of the probability it's like what's the probability I trade this system and I get blown out if I
45:51
start with a five thousand dollar account well that's good to know and
45:57
that's what that step does and it also there's a number I use in that that will
46:05
tell you if on a risk-adjusted basis this strategy is worth trading because
46:10
I've seen strategies that have made tons of money I mean I remember building
46:16
somewhere I thought oh this is the one that's just gonna put me into retirement
46:22
and then you start to trade it and you realize oh this had a lot of risk to it
46:27
a lot of drawdown and it it's not really worth trading you know it's so risky
46:34
so that helps you filter that out and then the last big test I do before then I I
46:42
try to put it in a portfolio and see how it fits and do the correlation and everything it's something I call
46:48
incubation which is really just watching a strategy
46:53
in real time uh it's really hard to do that because
47:00
everybody develops the system and they immediately want to go trade it
47:05
because wow look how great it looks look at all the money I'm gonna make I want to start trading this tomorrow
47:11
and what this incubation says is no don't do that wait
47:16
wait six months or even a year just put that strategy on a Shelf
47:23
and then come back and and check on it every few months or every month really hard to do but
47:31
uh I'd surveyed students who you know I asked him does this help you you know
47:36
because I always want to make sure what I'm teaching helps people and every time it's between 85 and 90 percent of
47:43
students say that step has definitely saved me money
47:49
um because what you end up doing if you've over optimized or done something wrong
47:56
the strategy usually falls apart in real time you know I I I used to
48:01
tell people how do you know if a strategy is good or not
48:07
take your back task take your Equity curve print it out on a eight and a half by eleven sheet of paper put it on the
48:14
wall you know 10 feet away from you okay and then
48:19
as you get some real-time performance for it add to that Equity curve and if
48:25
you can tell by looking at the shape of that Equity curve where you went live
48:31
that's probably a bad thing you know because a lot of times it'll be going up up up and then as soon as you start it
48:38
goes down well that's obvious but if you can't tell where
48:43
back test ended and real live trading began then that tells you you probably have a good system and that's what this
48:50
step is all about it's just about eliminating those common ones that are common
48:58
mistakes that people make and if you do that and it passes
49:04
are you guaranteed to make money no you know that they say past performance is
49:09
not indicative of future results for a reason but your chat your odds go way up when
49:18
you've one have a profitable back test and two have one that's worked in real time for
49:25
six months or so at least six months you'd be much better off and so once I'm
49:33
done with that step like I said I put it in the portfolio check correlation you
49:39
know you want to make sure it fits you don't want to trade two strategies that perform exactly the
49:44
same because you know it doesn't give you any diversification
49:50
and that's pretty much it I mean you know obviously the steps are pretty
49:55
simple it's putting it all together and then actually applying it that's the hard part you
50:02
know so many people will I don't need to test this this was some best-selling author who who wrote this great book so
50:10
I don't need to test his system it's got to work you wouldn't have made a book about it yeah he would have you know
50:18
it's uh or people yeah now it's more uh YouTube ads well I saw this YouTube ad
50:25
that this guy's making 200 a day every day uh you know and he's
50:32
only taking two trades or you know whatever you know making 50 a week
50:38
you got to watch out for that um you know the people one thing I've noticed they fall into the fallacy of
50:44
they see these outrageous claims 100 a week and they'll say well that's
50:51
on you know they'll know that that's not real but if I can get just a tenth of that
50:59
if I can get 10 a week that seems reasonable you know I'm not I'm not being greedy and going for 100
51:06
I'm just going for 10 percent I should be able to do that with this approach
51:11
and that's where people get messed up because yeah not only is this person lying about the 100 he's not making any
51:18
money with it it's losing weight probably and you do it you're not going to get 10
51:25
of it you're gonna get minus 10 of it you know so that's what you got to watch out for
51:31
because there's just so much misinformation out there it's it's ridiculous yeah it is unfortunate
51:39
um and you bring up a good point back test everything because people that you
51:44
think that are trading might not be right so that's why on this podcast I'm
51:50
really trying to vet who who I bring on you know with the track record because uh for obvious reasons I think it's
51:57
important now in regards to uh back testing uh what tooler and all that the
52:02
system Factory what tool are you using so I use uh tradestation now so I
52:08
mentioned I used to use Fortran now I use tradestation it's uh been around for
52:14
I want to say like 30 years or so so um some of the things are a little slow
52:19
a little dated um you know it's not necessarily ex state-of-the-art you know a lot of
52:25
people nowadays are using Python and um you know I think you can excuse me do
52:32
pretty well with that too um it's not necessarily the tool you use
52:37
but what you have to know is how to use it also how to break it so I always tell
52:43
people you know your trading software well enough if you know how to trick it
52:49
if you know how to create a great looking back test doing something wrong
52:55
then you probably know your software well enough so for example if you said hey create me a great system in
53:01
tradestation uh that's fake I could probably do it in about five minutes and
53:07
give you this beautiful Equity curve uh but for people starting out they
53:12
might create that Equity curve and think it's real so you've got to know your software well enough to know what's real
53:18
and fake but um yeah so I've been using tradestation for
53:24
you know almost 20 years and uh I haven't seen a need to jump to another
53:31
soft piece of software and and all the members in your community they're using tradestation also
53:39
really so the the process I teach works with any
53:45
back test software as long as you can optimize and do walk forward back testing but certain things like the club
53:52
that we have to standardize it everybody sends me code and trade station or I
53:58
help them develop the code in tradestation because I'm running everything on tradestation and I try to
54:04
make it uh consistent for everybody now people in Canada I had some students in Canada
54:11
for whatever reason the Canadian government uh has a lot of rules for Brokers and so tradestation for a while
54:18
they were just in Ontario and now they're you can't even use tradestation
54:24
Brokerage in Canada you can still use their software but you'd have to pay for
54:29
it which a lot of people don't like to do so those people tend to use multi-charts which
54:35
for the for people who don't know multi-charts is really it started off as uh at least what I heard was a Russian
54:43
hacker type knockoff of tradestation because tradestation used to be sold you
54:49
know for a certain amount of money and so these guys just programmed their own back test
54:56
software but use the same language that tradestation uses to develop strategies
55:01
so you can actually take a tradestation strategy and ninety percent of the time just plop it into multi-charts and
55:08
they'll work so multi-charts is a good tool too so I do have that I do use it I also have
55:14
ninjatrader and I use that not so much for development but um
55:20
you know there's a lot of good software out there that you can use so great so
55:25
anyone that would potentially want to join your membership Community you know as long as they're a systematic Trader
55:30
and maybe don't mind translating some things or getting some help with translating into trade stage and can
55:37
become a member yeah yeah I mean that's the the easiest way I mean you know the big thing I
55:43
always tell people there's a couple requirements one is you have to have some trading experience uh to benefit
55:50
from my class I don't teach people from the ground up uh it would just be
55:56
overwhelming for them to with all this material that I have and then two you have to know
56:03
a trading platform be familiar with it and as part of that so like 2A know how
56:09
to program some strategies in it and of course the platform you're using should be able to
56:15
back test optimize do walk forward testing do that kind of thing and as
56:21
long as you have that background uh you'd be pretty receptive to all the
56:28
material and all the ideas I give you that being said uh you know I offer as
56:36
per with my course uh tradestation help if you have problem
56:42
programming something since I know a tradestation I can help you but you know if you were a ninja Trader person I
56:49
can't help you with that code or metatrader because I don't know that code so it
56:55
ends up most most of my students end up either starting out with trade station
57:00
or moving over to trade station that makes sense perfect well we've only got
57:06
a few minutes left it felt like the conversation flew I guess a lot of great information
57:11
um you know is there any last piece of advice maybe something that is not well known or something that you feel is just
57:18
like you know what this is really critical for people to understand I call it the you know dropping Alpha right and
57:25
then finally how can people get a hold of you okay um I think the the biggest thing
57:31
nowadays it for people to realize is especially in trading education is to
57:38
realize how many of fraudsters and scammers are out there
57:43
it is incredible there's actually somebody who is offering a course to
57:51
people on creating your own trading course and
57:57
they sell you this I don't know how much thousand dollar package to make you a
58:02
trading Guru and in their information they actually say hey you don't have to be a Trader you don't have to be a good
58:07
Trader you don't even have to know anything and we'll make you so you're a good trading educator
58:13
I mean that's what it's come down to it it's nuts and uh so I always tell people
58:19
to live by the rule if it looks too good to be true it probably is so if you're looking at Twitter you're looking at
58:26
YouTube you're looking at anything or watching this and you think well what
58:32
are you saying is too good to be true go with that thought and run that to
58:37
ground because most things you're going to hear out there are false performance numbers and all kinds
58:45
of things and you see this with people getting uh arrested and getting sued by
58:52
the government for running scams so that's something I think people know it
58:59
so it's not necessarily a secret but people don't realize how prevalent it is and um
59:06
you know it's just so much out there it's it's uh it's tough and of course a
59:12
lot of people come to me and they say well hey I've been scammed before you're probably a scammer too
59:19
I'm like well you know what can I say except look at my stuff and draw your
59:24
own conclusion but uh yeah you know that you get lumped in there right but that's
59:30
the way it goes so for people who want to reach me if you go to my website
59:35
kgtradingsystems.com you'll see my contact information there you can also
59:40
email me directly I answer all my emails K Davy
59:45
kjtradingsystems.com uh I'm on Twitter KJ Trading
59:51
it's a good way to get a hold of me and uh you can also go to my YouTube channel uh I think it's called algo trading with
59:58
Kevin Davey or you can just in the search box of on YouTube just type in
1:00:03
Kevin Davey I think I thought I'd pop up near the top and uh yes there's a lot of different
1:00:09
ways to get a hold of me perfect and then I'll put all of those in in the show notes so wonderful all right Kevin I appreciate
1:00:16
your time today it's been an absolute pleasure uh definitely doesn't didn't disappoint and uh you know look forward
1:00:22
to uh to talking with you again sometime all right well great thanks for having me on and uh thanks to everyone who uh
1:00:29
listened to me ramble on for uh an hour or so all right take care of it all
1:00:35
right and that's it for alphacast episode four I'd like to thank you for tuning in you
1:00:41
can get all of the show notes and my personal takeaways at analyzingalpha.com front slash podcast
#Investing
#Commodities & Futures Trading
#Brokerages & Day Trading