Since I've had numerous questions asking me to show the results
of trading SSO in place of SPY for the Pro SPY Swing System, here's
some statistics and a video. Basically SSO results in about a
2X gain over SPY in most instances, sometimes more and sometimes
slightly less.
Also, there were questions about SSO degrading in place since
it's a leveraged ETF, however what's nice however is that on trades
which are held for a long time, such as 2 - 3 months, SSO does not
seem to degrade at all, in fact SSO seems to be giving a better
then 2X performance on those trades which were held for a long
time, see the video below
A leveraged ETF example Consider a hypothetical ETF that promises twice the return of
an index. Let's say you buy a share of the ETF for $100 while the
underlying index is at 10,000.
If the index goes up 10% the next day to 11,000, your
ETF should go up 20%, to $120. If the index goes from 11,000
back down to 10,000 the next day, that's a decline of 9.09%, which
means that the ETF should go down twice this much, or 18.18%.
A decline of 18.18% from the $120 price of the
ETF should leave it at $98.18. So even though the index ended
up right back where it started, the ETF is down 1.82%!
Why does this happen? Well, the ETF rebalances its assets
every day so that it will deliver the right multiple of the index's
returns
on that day. This means that if the index goes up
in value, the ETF will have to increase its exposure to the index
for the next day in order to get the right multiple, and if the
index goes down in value, the ETF will have to decrease its
exposure.
If the index keeps moving in the same direction every single
day, this should result in better-than-expected returns for
the longer period.
So a rigorous approach to the SSO vs. SDS choice creates two ETF
pairs and a new question has arisen.
Let's recap. Folks may desire leverage and seek alternatives to
SPY, but may also have good reason to avoid futures. Besides the
crazy higher risk with ES et al, there is less price detail for
opening, high, low, and closing price, since these vehicles trade
outside "regular" market hours. One should expect markedly
different performance for futures, so much different even that a
comparison may not be worthwhile. I'll not do it.
So on to the question. Whenever one receives a LONG signal, then
one either buys SSO or shorts SDS. Let's name this the LONG
pair.
And likewise, for a SHORT signal, one buys SDS or shorts SSO for
2X leverage. (SHORT pair)
Thus we have the interesting question to explore with ones
trusty spreadsheet: WHICH APPROACH RESPECTIVELY FROM THE TWO
DIRECTIONAL ETF PAIRS IS MOST EFFECTIVE?
The mathematical work is clear. One must segregate long and
short signals into their respective groups, all the longs together
and the shorts in a separate group. Next one matches the price data
for these actual trades from the back-test history to/with the
corresponding EFT pair and then simply calculates the best
performer.
It is not a given that the same approach will win for both
signal-types. The numbers will not lie.
More will be written on this later. For rest of today, I have to
prepare to give a science lecture for tomorrow. Feel free to begin
the work as it is laid out here on your own initiative(s).
THe biggest problem with using leveraged ETF's or Futures... is
that you really don't know how much you are risking...
For the sake of argument.. I think Matt said that the biggest
drawdown was 3 percent? This would mean 3k on
100,000 Keeping 97 percent of your capital "intact" on
a loss.. Do you know what the proper amount to invest in the
ETF or Futures contracts?
I personally trades 3 futures contracts when Matt said to go
long. I was underwater for awile... I added 3
futures contracts and then added 3 again until Matt
said to go flat. I was up over 14k at the end of
the flat signal. I found out that if was trying
to invest an equivalent of 100,000k was was putting way to
much at risk. I did get "lucky" with the trade in my
opinion...and will cut back the next time. There is
probably no easy way to figure out how much you should put into a
leveraged ETF or futures per
100k.
Does anyone know what the average percentage gain per trade
was?
For futures you look at the nominal value of the contracts you
trade. For ES it is $50 x Index point value, or as of Friday's
close, $68,175 per contract. In other words, in your example trade
above, your 9 contracts was the equivalent of about a $600K
position in SPY depending upon where the index was when you
entered.
Since the system doesn't use stops, you have to look at the
maximum losing trade to determine how much you are willing to risk.
Or, use the mid-trade maximum drawdown to determine how much heat
you could stand before bailing on the trade. In either case, the
first thing that you need to do is to determine for yourself what
the maximum loss to your account that you are willing to take on
any one trade.
ES example:
Let's say you have a $100K account and are willing to lose no
more than 3% ($3,000) on any one trade. If the system's historical
maximum losing trade was 3.36%, you would divide your $3,000 by
0.0336 (3.36%). The result is $89,286. That means that you could
comfortably trade only one ES contract (again, worth $68,175 right
now). When your account gets to $178.6K you could add another
contract. Or, you could add more contracts if your risk tolerance
is higher.
for simplicity... the 68,175 doesn't really
matter... The value would imply that you can lose
68,175 if the market went to zero. I
would say that a 20 point stop overnight would be good for a hugh
overnight lose. Hence 1000.00 per
contract. The nine contracts was an overload on
my part. The bigger problem is that
you can't by "fractions of contract" like you can by SPY
shares in odd numbers. The SPY also doesn't
evaporate in thin air like futures do. YOu can buy and hold
if you really get in trouble. I would
probably go long one futures contract in the future and add one at
each entry point... moving stops as necessary and putting my
initial stop at 1000.00 or 20 points... to simplify
things... On Matt's last trade that would have
made me a profit of a little under 5000. I
forget... but I think my overnight margin on futures is
around 3500 per contract.
If you're using stops then it's much easier: just figure out how
much you will lose if your stop gets hit. If you only want to lose
$1000 max, then you can just divide $1000 by $50 to get your ES
stop points value. $1000 / $50 =
I was talking about 20 points per contract.. ANyone who
would leave a 4 point stop overnight deserves to get stopped out or
shouldn't enter the trade in the first
place. Don't
forget... there is more gaps up overnight than gaps down.
Posted by marketguy on 30th of Apr 2011 at 09:07 pm
the Dow is going to "0", no wait maybe to "30,000" but then gold
is going to $10,000 an ounce and silver is going to $1,000 an
ounce so what does the Dow matter? on the other hand, maybe gold is
going to $500 again and silver is going back to $8....but then
where does that leave the Dow? Gas at $4 today in my neck of
the woods but yet the local shopping mall was packed and I mean
PACKED today and restaurants have been busy even on Tuesdays and
Wednesdays.....Japan earthquakes, PIGS blow ups, Syria shooting
everyone in sight, tornados, floods and a Fed with a blank check to
do whatever they want....
now that it's clear as a bell, what should I be doing on Monday
again?
but why bother shorting the inverse funds for the SPY system
when matt substituted the trades -- and cubby did also -- and SSO
works as well or better than SPY? IF you want that additional
exposure, of course.
the question was whether you could pick up some extra return by
shorting SDS to go long or SSO to go short in order to benefit from
leveraged ETF tracking errors over time. However as per the earlier
post maybe since the system is "right" so often the ETFs will move
in the right way and tracking error will not be an issue
Posted by stevedfw on 28th of Apr 2011 at 12:11 pm
Just threw that out for discussion. So what is being proposed is
to use SSO for long and short trades and NOT using SDS at all
(I.e.for short signals), correct?
but test it first -- I was surprised when I looked at a long
term trade shorting FAZ vs buying FAS .... can probably find my
post on it if you search if interested
we can do these comaprisons ourselves -- matt just got to it
before I did it this weekend.......... cubby on the blog did
it also........... Just take the trade list and go back trade
by trade and substitute the levereged fund of your choice on a
spreadsheet.....................
I have never traded futures but have just signed up to do so,
therefore I'd be grateful if someone could correct me if I am wrong
in the following assertion:
- Regarding the E-mini ES, the contract was worth much less when
SPX was at the March 09 lows around 666 vs today at 1350ish. As
such, I believe I am correct in saying a 1% move in the ES in March
09 resulted in less points (and therefore $$) profit/loss per
contract vs an equivalent 1% today.
- In other words, although using Matt's SPY system to trade the
ES will result in the same winning and losing trades, the actual
profits and losses in $$ will differ to a fixed $100,000 per SPY
trade because of the varying values of the ES contract over
time.
SPY System - substituting SSO for SPY
Posted by matt on 28th of Apr 2011 at 12:09 am
Since I've had numerous questions asking me to show the results of trading SSO in place of SPY for the Pro SPY Swing System, here's some statistics and a video. Basically SSO results in about a 2X gain over SPY in most instances, sometimes more and sometimes slightly less.
Also, there were questions about SSO degrading in place since it's a leveraged ETF, however what's nice however is that on trades which are held for a long time, such as 2 - 3 months, SSO does not seem to degrade at all, in fact SSO seems to be giving a better then 2X performance on those trades which were held for a long time, see the video below
http://breakpointtrades.com/jing/2011-04-27_2301.swf
as long as the direction is correct
Posted by rikkwan on 28th of Apr 2011 at 10:38 am
Leveraged ETFs don't hurt you if you get the direction of the move "right."
correct, I saw no time
Posted by matt on 28th of Apr 2011 at 10:40 am
correct, I saw no time delay. However again your losses and draw down will also be 2X as large, not just your gains.
this educational article from Schwab about leveraged ETFs summarizes the issue nicely
Posted by rikkwan on 28th of Apr 2011 at 11:10 am
http://www.schwab.com/public/schwab/research_strategies/market_insight/investing_strategies/exchange_traded_funds/leveraged_and_inverse_etfs_not_right_for_everyone.html
"If the index keeps moving in the same direction every single day, this should result in better-than-expected returns for the longer period."
http://www.schwab.com/public/schwab/research_strategies/market_insight/investing_strategies/exchange_traded_funds/leveraged_and_inverse_etfs_not_right_for_everyone.html
A leveraged ETF example
Consider a hypothetical ETF that promises twice the return of an index. Let's say you buy a share of the ETF for $100 while the underlying index is at 10,000.
If the index goes up 10% the next day to 11,000, your ETF should go up 20%, to $120. If the index goes from 11,000 back down to 10,000 the next day, that's a decline of 9.09%, which means that the ETF should go down twice this much, or 18.18%.
A decline of 18.18% from the $120 price of the ETF should leave it at $98.18. So even though the index ended up right back where it started, the ETF is down 1.82%!
Why does this happen? Well, the ETF rebalances its assets every day so that it will deliver the right multiple of the index's returns on that day. This means that if the index goes up in value, the ETF will have to increase its exposure to the index for the next day in order to get the right multiple, and if the index goes down in value, the ETF will have to decrease its exposure.
If the index keeps moving in the same direction every single day, this should result in better-than-expected returns for the longer period.
rikkwan and ETF rebalancing
Posted by cubby on 30th of Apr 2011 at 05:41 pm
So a rigorous approach to the SSO vs. SDS choice creates two ETF pairs and a new question has arisen.
Let's recap. Folks may desire leverage and seek alternatives to SPY, but may also have good reason to avoid futures. Besides the crazy higher risk with ES et al, there is less price detail for opening, high, low, and closing price, since these vehicles trade outside "regular" market hours. One should expect markedly different performance for futures, so much different even that a comparison may not be worthwhile. I'll not do it.
So on to the question. Whenever one receives a LONG signal, then one either buys SSO or shorts SDS. Let's name this the LONG pair.
And likewise, for a SHORT signal, one buys SDS or shorts SSO for 2X leverage. (SHORT pair)
Thus we have the interesting question to explore with ones trusty spreadsheet: WHICH APPROACH RESPECTIVELY FROM THE TWO DIRECTIONAL ETF PAIRS IS MOST EFFECTIVE?
The mathematical work is clear. One must segregate long and short signals into their respective groups, all the longs together and the shorts in a separate group. Next one matches the price data for these actual trades from the back-test history to/with the corresponding EFT pair and then simply calculates the best performer.
It is not a given that the same approach will win for both signal-types. The numbers will not lie.
More will be written on this later. For rest of today, I have to prepare to give a science lecture for tomorrow. Feel free to begin the work as it is laid out here on your own initiative(s).
Best regards, Cubby
Leverage
Posted by zach06 on 30th of Apr 2011 at 06:22 pm
THe biggest problem with using leveraged ETF's or Futures... is that you really don't know how much you are risking... For the sake of argument.. I think Matt said that the biggest drawdown was 3 percent? This would mean 3k on 100,000 Keeping 97 percent of your capital "intact" on a loss.. Do you know what the proper amount to invest in the ETF or Futures contracts?
I personally trades 3 futures contracts when Matt said to go long. I was underwater for awile... I added 3 futures contracts and then added 3 again until Matt said to go flat. I was up over 14k at the end of the flat signal. I found out that if was trying to invest an equivalent of 100,000k was was putting way to much at risk. I did get "lucky" with the trade in my opinion...and will cut back the next time. There is probably no easy way to figure out how much you should put into a leveraged ETF or futures per 100k.
Does anyone know what the average percentage gain per trade was?
Calculating Risk Using Futures
Posted by pdquig on 30th of Apr 2011 at 08:05 pm
For futures you look at the nominal value of the contracts you trade. For ES it is $50 x Index point value, or as of Friday's close, $68,175 per contract. In other words, in your example trade above, your 9 contracts was the equivalent of about a $600K position in SPY depending upon where the index was when you entered.
Since the system doesn't use stops, you have to look at the maximum losing trade to determine how much you are willing to risk. Or, use the mid-trade maximum drawdown to determine how much heat you could stand before bailing on the trade. In either case, the first thing that you need to do is to determine for yourself what the maximum loss to your account that you are willing to take on any one trade.
ES example:
Let's say you have a $100K account and are willing to lose no more than 3% ($3,000) on any one trade. If the system's historical maximum losing trade was 3.36%, you would divide your $3,000 by 0.0336 (3.36%). The result is $89,286. That means that you could comfortably trade only one ES contract (again, worth $68,175 right now). When your account gets to $178.6K you could add another contract. Or, you could add more contracts if your risk tolerance is higher.
Make sense?
ok..but....
Posted by zach06 on 30th of Apr 2011 at 08:58 pm
for simplicity... the 68,175 doesn't really matter... The value would imply that you can lose 68,175 if the market went to zero. I would say that a 20 point stop overnight would be good for a hugh overnight lose. Hence 1000.00 per contract. The nine contracts was an overload on my part. The bigger problem is that you can't by "fractions of contract" like you can by SPY shares in odd numbers. The SPY also doesn't evaporate in thin air like futures do. YOu can buy and hold if you really get in trouble. I would probably go long one futures contract in the future and add one at each entry point... moving stops as necessary and putting my initial stop at 1000.00 or 20 points... to simplify things... On Matt's last trade that would have made me a profit of a little under 5000. I forget... but I think my overnight margin on futures is around 3500 per contract.
The other way of looking at it...
Posted by pdquig on 30th of Apr 2011 at 10:19 pm
If you're using stops then it's much easier: just figure out how much you will lose if your stop gets hit. If you only want to lose $1000 max, then you can just divide $1000 by $50 to get your ES stop points value. $1000 / $50 =
20 ES points for one contract
10 ES points for two contracts
4 ES points for five contracts, etc.
not really.
Posted by zach06 on 30th of Apr 2011 at 10:42 pm
I was talking about 20 points per contract.. ANyone who would leave a 4 point stop overnight deserves to get stopped out or shouldn't enter the trade in the first place. Don't forget... there is more gaps up overnight than gaps down.
I think....
Posted by marketguy on 30th of Apr 2011 at 09:07 pm
the Dow is going to "0", no wait maybe to "30,000" but then gold is going to $10,000 an ounce and silver is going to $1,000 an ounce so what does the Dow matter? on the other hand, maybe gold is going to $500 again and silver is going back to $8....but then where does that leave the Dow? Gas at $4 today in my neck of the woods but yet the local shopping mall was packed and I mean PACKED today and restaurants have been busy even on Tuesdays and Wednesdays.....Japan earthquakes, PIGS blow ups, Syria shooting everyone in sight, tornados, floods and a Fed with a blank check to do whatever they want....
now that it's clear as a bell, what should I be doing on Monday again?
Statistics
Posted by cubby on 30th of Apr 2011 at 06:40 pm
Asked by a member: "Does anyone know what the average percentage gain per trade was?"
Yes, that may be found on the SPY FAQ page.
Maybe the safest way if
Posted by stevedfw on 28th of Apr 2011 at 11:19 am
Maybe the safest way if you are going to use SSO and SDS is to only short them with the SPY signals.
but why bother shorting the
Posted by Michael on 28th of Apr 2011 at 12:00 pm
but why bother shorting the inverse funds for the SPY system when matt substituted the trades -- and cubby did also -- and SSO works as well or better than SPY? IF you want that additional exposure, of course.
the question was whether you
Posted by bkout3 on 28th of Apr 2011 at 12:16 pm
the question was whether you could pick up some extra return by shorting SDS to go long or SSO to go short in order to benefit from leveraged ETF tracking errors over time. However as per the earlier post maybe since the system is "right" so often the ETFs will move in the right way and tracking error will not be an issue
Just threw that out for
Posted by stevedfw on 28th of Apr 2011 at 12:11 pm
Just threw that out for discussion. So what is being proposed is to use SSO for long and short trades and NOT using SDS at all (I.e.for short signals), correct?
but test it first --
Posted by bkout3 on 28th of Apr 2011 at 11:45 am
but test it first -- I was surprised when I looked at a long term trade shorting FAZ vs buying FAS .... can probably find my post on it if you search if interested
I assume that if we
Posted by ditch on 28th of Apr 2011 at 09:53 am
I assume that if we can substitute SSO that we can therefore short via SDS ?
ditch - you would think
Posted by Michael on 28th of Apr 2011 at 09:59 am
ditch - you would think so, and probably, but you'd have to go back and list the trades to know for certain.
Thanks, Matt. Now, if you have
Posted by algyros on 28th of Apr 2011 at 09:41 am
Thanks, Matt.
Now, if you have the time, it would be interesting to see what the results would be using BXUB, which is the 3X SPX etf.
we can do these comaprisons
Posted by Michael on 28th of Apr 2011 at 09:44 am
we can do these comaprisons ourselves -- matt just got to it before I did it this weekend.......... cubby on the blog did it also........... Just take the trade list and go back trade by trade and substitute the levereged fund of your choice on a spreadsheet.....................
Great stuff as always Matt,
Posted by aharon on 28th of Apr 2011 at 09:35 am
Great stuff as always Matt, much appreciated. I'm curious if the same holds true for the ES mini as well.
Title: Using SPY system signals
Posted by philosoraptor on 28th of Apr 2011 at 10:17 am
I have never traded futures but have just signed up to do so, therefore I'd be grateful if someone could correct me if I am wrong in the following assertion:
- Regarding the E-mini ES, the contract was worth much less when SPX was at the March 09 lows around 666 vs today at 1350ish. As such, I believe I am correct in saying a 1% move in the ES in March 09 resulted in less points (and therefore $$) profit/loss per contract vs an equivalent 1% today.
- In other words, although using Matt's SPY system to trade the ES will result in the same winning and losing trades, the actual profits and losses in $$ will differ to a fixed $100,000 per SPY trade because of the varying values of the ES contract over time.
Each E-Mini S&P point is
Posted by jambo1 on 28th of Apr 2011 at 10:19 am
Each E-Mini S&P point is $50 thus each tick is $12.5 and has to be worked out as such.
Much appreciated. Thanks, Matt.
Posted by frtaylor on 28th of Apr 2011 at 09:14 am
Much appreciated. Thanks, Matt.
thanks so much for doing
Posted by Michael on 28th of Apr 2011 at 08:09 am
thanks so much for doing the SSO comparison.
So how will the SPY
Posted by ranger on 28th of Apr 2011 at 07:36 am
So how will the SPY system be priced?
In physical silver...
Posted by lachasse on 28th of Apr 2011 at 07:47 am
In physical silver...
Thanks for the update Matt.
Posted by michelem on 28th of Apr 2011 at 01:42 am
Thanks for the update Matt. Sorry to hear about funeral...