I'm just learning about the UltraShort ETFs which seeks investment results that correspond to twice (200%) the inverse (opposite) the underlying index.

If the index increases 50% won't the ETF become worthless? How can a fund exist which is 'expected' to be worthless someday in the future?


Any good articles that would help me understand these?

http://www.proshares.com/funds/sdd.html

Doesn't quite work that way because they are constantly resizing their positions.

point me towards one time in the history of any market ever where the market index rose 50% in one day.

its more likely that ultra long will "go bust" according to you b/c markets are far more likely to fall precipitously than to rise so.

the ultrashares funds aim for double the *daily* returns of the underlying index. they don't deliver double the logn term returns.

i'm not 100% sure i understand everything about them other than they don't perform like they say they will (i.e. a lot of hidden costs, one of which is that they don't return double the goal of whatever, and sometimes may not even provide double the goal of whatever on a day to day basis), so stay away from them is the only advice i can give.

Barron

Barron,

I was just reading this thread,and you made a comment about 'your reco to short' the S&P in yesterdays trading thread to the OP.You obviously just edited/erased it(unless my eyes just decieved me)?

Anyway you posted as follows:

2:48 i'd get short the S&P500 at 1350.86 (-20% signal)
Barron

My question is WHY/HOW did you come up with that number,and what does a -20% signal mean?

stephen,

i think you should take a break from the screens. that quote you put up is still in the thursday trading thread. i never deleted it, nobody touched it (i checked the thread moderation application and there is/was zero action regarding that post).

i'll asnwer your question ehre though:

that specific number is an estimate (i.e. roundabout guess). the "signal" is how strongly convinced you are regarding the trade. it ranges from -100% (most bearish) to +100% (most bullish)

the % is of the maximum position you would take in that trade if you had the strongest possible conviction you had. the size of the maximum position is a function of the volatility of the position on its own and its correlation with the rest of your actively traded portfolio.

while i don't have a system set up to analyze that, i still use the terminology because i find it very good for quantifying (qualifying in this case) how much you'd risk relative to the total amt you'd put behind the trade ... i.e. "firing all bullets" as yowsers put it.

if my signal was 100% bearish, i'd "fire all bullets" (allotted to that trade).

Barron

i dont understand them either but i made more off them in january this year then i did all of last year going long normal stocks, there are definetely fees and it isnt exactly x2, but they are working in this kinda market

Yeah my main question about the Ultrashorts is how much will they be worth in 10 years, assuming the market does its normal 10%/yr gain.

Barron

I apologize about the deletion comment above(you had referred to it in another thread about shorting to another OP....sorry my bad).And your probably right about tooo much 'screen exposure' for me this week

I understand your short entry/price level......but am still confused as to your rating system/percentage and how you devised these #'s?

Are they figures YOU have tried/used before when entering a position?

Stephen

a simple way to describe it is in Z terms.

you have a projected/estimated value/level for your trade and the market is at some other level.

let's say you're trading a 10yr bond. it is currently priced at P and your projected/estimated price is P*

(P-P*)/ std(P) is the Zscore where std(P) ist he standard deviation (or some measure of standardized risk) over the relavent history of P's price movements.

this tells you how far your projected price is from the current market price in terms of some standardized measure of risk.

as Z gets bigger, the distance between P* and P in st.dev terms gets bigger and your conviction (assuming your hypothesis is solid and you trust it) similarly increases. thus your signal should start approaching 100%.

the function that maps Z to the signal should be a logistic function (S curve) where it is very hard to get near +/- 100%, but it doesn't take a huge distance from 0 for Z to trigger the signal to move north or south of 0%.

obviously here, you should have a projected price and the market price and the relavent history of the volatility of that price.

in this case (the -20% signal), i basically used it to give a jist of how strong i was convinced of the trade. i think that the S&P500 is overvalued by some portion of a standard deviation.

Barron

It sounds more or less like a mathematical version of using 'Bollinger Bands' to trade at market extremes?

How accurate/successful do you find your signals?

TTYL,
Stephen

they aren't "my signals" so to speak. the fund i worked for constructed their entire active portfolio this way with hundreds of thousands of indicators feeding some 500 signals. this methodology has been exceedingly successful over 16 years of trading in the most liquid markets in the world.

they now run over $30bil in their main hedge fund vehicle.

while there i used much smaller signal generators to do a guesswork kind of alpha portfolio that they let us trade (off of our bonus) and i did very well with it.

the methodology sounds more mathematical than it is. the real meat of it is logically and systematically determining the right indicators that build the "projected price."

Barron

Sounds pretty intersesting,and if you dont mind try and post a few buy/sell signals...... lets say for just the S&P,so I can compare them to my work

You can even PM them to me if you dont want to post them here

Thanks
Stephen

Right, how can the management company run a fund where the assets they are investing in (those that are targeted to -2x the market) are likely to slowly become worthless. Then when you add the fees on top of the normal returns it seems like they are even more likely to slowely become worthless.

Please help me wrap my mind around these instruments.

You can't predict exact long term returns because it depends on the specifics of daily returns. But the short answer to why they don't go busto is that they resize positions every day. If the market goes against them they scale back.

I don't think they scale back if the market goes against them. The goal of the instrument is to perform -2x the market, good or bad.

They must work like this:
EFT's are like mutual funds that trade as a price consistent with a share of the value underlying fund.
http://www.mutualfundsindia.com/etf.asp This article explains that market makers invest in creation units which are then traded on the exchange. You could theoretically take your shares to the fund house and redeem them.

So let's say you buy 2 shares at $100 today and the stock market increases 25%, now the value of each share should be near $50. But there are more people who want to invest in the fund, so I doubt that they would want the price of the shares near zero. So they do a reverse split and double the price and cut in half the number of shares. So now you own 1 share at $100.

The only thing that will keep sufficient capital in these funds over the long run is the market makers investing more money as a hedge.
Correct analysis?

This is from the proshares page. How is this possible?

"Have gains possibly qualify for long-term capital gains tax treatment? Yes"

http://www.proshares.com/abtfunds/AboutShortProShares

You're wrong. I'm not sure why you keep disagreeing with the answers that several people have provided. They are aimimg for a -2 DAILY correlation with the market and adjust positions accordingly each day.

I'm not trying to disagree, it's just that I don't understand the difference you seem to be talking about.

How can they achive a daily correlation each day without having the weekly or yearly correlation be the same as well?

I'm sure others on here can explain this a lot better than I can, but here's a simple example to illustrate the difference. If someone can check my math I'd appreciate it.

You invest $1,000 in a Nasdaq ultra-short fund with QQQ's current price at 50. They short 40 shares of QQQ to get you the target -200% return vs. QQQ. The first day you hold it, QQQ goes up 25% to $62.50. Your position is now worth only $500. If they did it your way, they would continue to hold a position of short 40 shares for you at the $62.50 price. For every 1% QQQ went up ($.625), you would lose $25...which is 5% of your position. That's a lot more than the 2X leverage your short position started out with! So what they'd actually do after that first day is reduce the position size to somewhere around 16 shares, so that another 25% increase tomorrow would again result in a 50% loss for you. So your way, back to back 25% increases leaves you with $0. My way (which approximates what they actually do) it leaves you with $250. The difference between the two is even greater if the losses are more gradual.

this is pretty much right. there are articles that relay this same info but zoo is dead on in terms of what he means by "resizing"

in order to get the same % return, they can't hold a static position size. this leads to deviations of longer term performance. my goal is to create a leveraged portfolio, however, proshares 2x funds are not suitable for this purpose since i need things that will acheive 2x LONG TERM performance (which proshares doesn't)

Barron

Ok, now I understand. Thanks for taking time to explain it.

I invest in an ETF called SDS. It is an ultrashort ETF. It has served me well over the past three months. If you have any other questions about it, let me know.