July 19, 2007
Stock Du Jour (HUM) & Random Observations
Selling from the get-go, continued all day with a little perk up in the last half hour.
Notable New Lows: HUGE number of “Bancorps” including Popular (BPOP) which when I look at their numbers just looks dirt cheap to me — there’s so much “deep value” in this sector now, a lot of people must be licking their lips and building monster long positions; Interactive Brokers (IBKR); Circuit City (CC); homebuilders Lennar (LEN) and Beazer (BZH); and Wally Weitz’s favorite, Redwood Trust (RWT).
Notable New Highs: Exxon (XOM), CSX (CSX) and Union Pacific (UNP) and Canadian National (CNI) and Canadian Pacific (CP), St. Jude (STJ) and Humana (HUM).
I’ll feature a monthly chart of Humana (HUM)… recall that at the height of the dot com bubble there were a lot of good opportunities in deeply out of favor companies.
July 20th, 2007 at 12:11 am
Chairman,
Can you help settle an argument for me?
It is my hypothesis that the QID can go to zero. I actually called the Proshares office the other day, and spoke with a young man, who said in response to my thesis….”That’s a very interesting question. I’ll have to look into that…”
Now the NDX would have to rise 30% from here, so I’m not saying it will happen overnight…..
Thanks for any light you can shed on this subject!
July 20th, 2007 at 10:09 am
Linda: I don’t think it can go to zero but it can approach zero … if the QQQQ went up 30% from here the QID would drop 60% to $16 and change, if the Q’s rose another 30% and the QID’s drop 60% again they’d be at $6 and change; you’d be approaching zero but never get there… see what I mean?
July 20th, 2007 at 1:29 pm
To confirm Maoxian’s explanation. The prospectus states that the goal is to achieve the inverse of twice the DAILY PERCENTAGE move.
One key here is that the target move is based on the percentage change and not the absolute dollar change.
The other key point is that the goal is based on the daily change and not the long term change.
Thus, as the chairman explained, the cubes could go up 5% a day for ever and the QID will lose roughly 10% of its starting daily value every day, never hitting $0.
July 20th, 2007 at 2:58 pm
I recall having see a related question about QID, perhaps on maoxian.com many months ago: can you short QID, and mightn’t that be a good thing to do if one were correctly (if it turned out that way) bullish on NDX (as opposed to going long QLD)?
Through doofus bungling, I can answer the first question: yes. I accidentally did it one day. At least I think I did it. I meant to go long QID, and so I immediately reversed the trade. (I didn’t dwell on it, especially since I was embarrassed to have profited by doing the opposite of what I intended, but I’m pretty sure that that is what happened.)
I am not however sure of the answer to the second part of the question. Anyone?
July 20th, 2007 at 6:53 pm
Thank you Chairman and Born2Code! I get it now…wish the folks at Proshares could have been so succinct.
July 20th, 2007 at 9:36 pm
Expect a “reverse split” if the QID gets too low in price.
Going short one product may incur margin fees that going long might not incur, depending on timeframe held and other variables. Also can’t go short in some accounts. Otherwise, yep, it’s about the same thing, other than possible tracking error of the instrument.
July 20th, 2007 at 10:46 pm
There is another problem with shorting QID. It pays a dividend. I played QID as a swing trade sometime after the Feb. 27th event and was surprised a while later when i got a dividend credit in my account. I only held it for few days but my timing was such that i held thru the ex- date and got paid.
They pay a dividend because they use futures to achieve the double inverse and they put up cash as margin requirements which earns interest.
Any how, if you short it, then you have to pay the dividend back to the original owner which would add an extra layer of complexity that you could do without simply by going long QLD instead.
July 20th, 2007 at 11:00 pm
Born2: If you were simply day trading the QID you would never face that problem.
July 20th, 2007 at 11:04 pm
So the big question is what would happen if the NDX went up 50% or more in a single day??? How do you lose more than 100% w/o going at least to zero?
July 20th, 2007 at 11:10 pm
Jerry: I wouldn’t hold my breath for that one day 50% rise. ;-)
July 20th, 2007 at 11:13 pm
If NDX goes up 50%, the change is 1.50. Invert that to 1/1.50 = 0.667, double the change (-0.333 to -0.666), and the QID goes down 66.6%.
Similarly, if the NDX goes down 33.5%, the change is 0.667. Invert that to 1/0.667 = 1.50, double the change (+0.50 to +1.00), and the QID goes up 100%.
July 20th, 2007 at 11:17 pm
Bill: Thanks for doing the math; it’s late and my brain is tired.
July 21st, 2007 at 12:34 am
Thanks for clarifying Bill. Two interesting things emerge from your math. First, the QID can still go to zero if the NDX doubles in one day, theoretically. [1/2.0=0.5; double the change and -0.5 becomes -1.0] It seems the Chairman and Born2Code’s shortcut of multiplying and changing the sign were just a linear approximation that works for very small changes in NDX. Second, plotting the true relationship shows a concavity property that only adds to QID’s appeal over shorting — that is, it delivers bigger profits for a given down move in NDX than it does losses on an equal magnitude up move. For example, if the NDX moves down 10% in a day, QID is up 22.2%, BUT if NDX moves up 10% in one day, QID only loses 18.2%!
July 21st, 2007 at 11:18 am
What about the management company’s fees? They’re in the mix. The fees are small, because this is an index ETF, but they’re there. Do you get an advantage by shorting QID that way (instead of going long QLD)? Do the fees not then work for you instead of against you?
Also, am I right that dividends don’t work against shorts any more than they directly benefit longs— because the price of the security drops the moment it goes ex-dividend so it all comes out in the wash anyway?
July 21st, 2007 at 12:13 pm
I decided to take a stab at answering my question using some real data on the QID/QLD YTD returns.
Using Yahoo’s historical data, I find that the split- (there were none) and dividend- (there were some) adjusted return ratios for QID and QLD are respectively 0.78 and 1.29 .
So, if I had put up a dollar to go long QLD I now have $1.29 .
Now, if I had put up a dollar to sell short a dollar’s worth of QID (no loan from the broker involved, but a margin account required), I would now have that dollar plus the broker’s sweep interest on it (about 5% per year at IB), namely about $1.028 by now, plus the $1 that the broker got for the stock (on which no interest is paid) minus the ongoing obligation to buy it back, with the current price being $0.78 .
So in all, the current YTD return ratio for going long QLD would be 1.29, and the current YTD return ratio for going short QID would be (1.028 + 1.00 - 0.78)/1, or about 1.25 .
So, using real data, shorting QID would not have been the way to go. But it is the case that QID and QLD don’t perfectly track NDX, and thus they don’t track each other perfectly anyway. That could be the main consideration.
July 21st, 2007 at 7:54 pm
Mike: I was referring to a benefit of *buying* the QID over any actual shorting strategy of the *index*, because the concave relationship juices returns if you are right, and dampens losses if you are wrong. I wasn’t referring to shorting the QID over going long QLD…but that was an interesting comparison of those strategies.