Friday 25 February 2011

Week of 2011 FE 25

*Whew* Boy!  What a week!  I lost 0.1%, while SPY lost 1.8%.  Basically, I lost 1½ weeks’ profit, while SPY lost 2½ weeks.

% gain, less 0.04%/day: close stop Realized
Symb Buy Fri   Tue   Wed   Thu   Fri gain
 
DRETF DE16 4.5 3.9 3.8 1.6 4.3
CAT FE07 3.7 1.7 2.3 1.8 1.9 2.2 2.9%
SPY FE08 0.5 -0.2 -0.5 -0.9 -0.9 -0.5 0.3%
TNA FE09 4.2 1.4 0.1 -2.6 -1.7 1.5 3.3%
LINTA FE17 0.6 -2.5 -2.7 -4.5 -4.6 -2.9 −2.3%
CR FE18 0.1 -2.5 -3.3 -5.7 -6.6 -4.3 −2.5%
MMS FE18 2.0 -3.1 0.8 -0.5 0.0 -0.5 0.2 -0.5 0.6 -0.6
ARE FE18 -0.8 -1.8 -1.6 -3.0 -4.5 -2.7 −1.7%
PROV FE23 -0.9 -1.7 -1.5 0.3 −1.8%
PIKE FE23 -5.9 -9.9 -3.3 -8.1 -2.0 -8.1
CXO FE23 0.4 -4.6 -0.8 -3.8 1.2 -3.1
CBS FE25 5.3 -3.8
LRN FE25 1.8 -4.9
(Average) −0.4%

Monday: President’s Day.  US markets closed.  Canadian markets also closed, for Family Day.

Tuesday:  My emergency stops worked!  Losses for the day were capped at 1.3%, which happened to be the same as SPY’s loss.  I sold ¼ CAT at the open, then the final quarter was stopped out.  I sold ½ MMS at the open—it was the only sell-stop that didn't trigger, so I still have the other half.  Some settled cash came in, so I bought CXO (energy), PIKE (energy construction) and PROV (microcap bank) because they didn’t drop when all the other stocks did.

Wednesday: By the time I got up in the morning, it was too late to place a 4% stop on PIKE, which had already dropped 6.1% since the open!  So I gave it a 10% stop and hoped.  Later it rose to -5.9% for the day.  CXO rose nicely but PROV fell and they cancelled out.  PIKE is about 10% of the account so its 6% loss made me lose 0.6% overall for the day.  Still, SPY did even worse and lost another 1.3% for the day.

Thursday: PIKE recovered half its loss.  PROV fell a little more and got stopped out.  I had money to invest, but nothing looked good, so I bought more CXO at 1:30pm which is usually its low point of the day.  But today that was the high!  CXO ended its day where it had begun.  DRETF came out with its annual report, which was very good but apparently not good enough, so the stock dropped its recent gains.  I lost 0.3% for the day, while SPY held to neutral.

Friday: CXO rose, and it is ⅓ of my entire account!  PIKE and MMS rose a bit, while DRETF recovered its losses for the week.  New purchase LRN rose nicely.  New star CBS rose 5% on its first day!  I’ve put in an order to sell half of it on Monday.  Overall I gained 2% for the day, while SPY gained only 1%.

Part of the problem with DRETF is that I was checking the daily-close values too early.  As a non-listed security, its trading day doesn’t end until 6:30pm.

The new “Active Trader” platform

So I got Schwab’s “Active Trader” thing (which was previously StreetSmart.com).  It’s got good points and bad points.  It does offer trade-stops, but only as “conditional alerts” which are treated separately from trades (when the alert fires, it creates a trade order).

Schwab recommends that trade-stops be used only for NYSE or AMEX stocks, as they are unreliable for NASDAQ or PINK.  Every time you log in, it laboriously lists each active conditional alert in a popup window.  Clearly, they don’t want you to use too many!  Judging from all the alerts they put up, it seems the two things Schwab is afraid of are that you will ask them to reverse a trade “because I didn’t intend to do that!” or that you will walk away in a snit because your stops sold for much less than their trigger prices.

The “bracket” orders work well.  You can enter a sell-stop at the same time as the initial purchase, even though you don’t have the shares yet.  (Too bad they always trigger on bid prices.)  You can enter a sell-limit and a sell-stop on the same shares, which I’m using to sell ½ of CXO and LRN if they go up a little more.  In the plain system, multiple orders for the same shares are rejected; in the active-trader system, you get a warning that the stop will sell a reduced number of shares if the limit triggers first.  Why yes, that *is* what I meant!

New rule: When the market drops a lot, don’t buy anything the next day.  I made this same mistake in both of the last two corrections.  Insteading of buying something that didn’t drop when everything else did, wait another day and buy things that didn’t drop either on the Big Day or the day after.  This would avoid problems like PIKE that didn’t drop *yet* by Tuesday’s close but then reality caught up with it on Wednesday.

For Monday: pick another stock to buy, to use up the cash and the PROV money that will come in on Tuesday.

Friday 18 February 2011

Week of 2011 FE 18

My stock-trading results were not so hot for this week: up only 0.3%, while SPY was up 1.1%.  The two main reasons seem to be ADTN (lost 0.4% just on that one stock) and the cooling down this week of stocks I've held since last week (especially ATML).  With certain notable exceptions such as ADTN and BAGL, the stocks I pick do seem to jump soon after I buy them, so I guess the problem remains that I hold them too long and their fire dies out before I sell.

Still, I’ve made 2.5% over the last four weeks (the green bars), which would be 32% per year!  SPY is up 4.8% (the red line) which would be 62%; obviously neither of these annualized figures is believable because the current rate of growth cannot be sustained.  I think it will end when the Fed stops their Quantitative Easing program (maybe in June).  My ups and downs don’t seem to be correlated with SPY’s, suggesting that my problems are poor execution rather than just the market’s fickle mood.  That’s good news if true because my execution can be improved but the market’s mood can’t.

Stock
Symb
Hold
Days
Week
Chg
Gain in ‱/day: close stop
Fri   Mon   Tue   Wed   Thu   Fri Actual
 
DRETF 45⁺ 4.6% 5 5 7 9 9 14
ETN -1.4% 40 8 29 12 11 22 16 15 12
GIII -1.2% 7 -23 0 -24 -11 -10 -4 -12
ATML -2.7% 96 0 100 24 61 46 49 42 41 41 40
MKL 0.9% 18 -4 10 0 6 4 8 4 10 11 4
CAT 10⁺ 2.2% 53 -14 43 -12 31 11 31 14 27 16 35 24
SPY 9⁺ 1.1% 7 -14 10 -12 4 -7 11 -3 9 2 9 3
TNA 8⁺ 4.6% 55 -70 71 -56 31 -2 42 5 50 11 46 17
ADTN -0.8% -40 -103 -59 -82 -47 -68 -3 -15 -62
ENTG 2.7% 247 -147 98 -2 77 20 66 40 48 59
LINTA 2⁺ 4.5% 33 -80 20 -47
CR 1⁺ 3.1% 10 -59
MMS 1⁺ 5.3% 73 -12
ARE 1⁺ 1.0% -22 -42

I’ve been using this style of daily-progress table for a few weeks now, but I’m starting to lose enthusiasm for it.  It seems unable to answer important questions, like “How come I didn’t make much money this week?”  The purpose of focusing on gain-per-day figures is that “time is the enemy”, so a stock that gains 5% in a week is twice as good as one that gains 5% in two weeks.  But a stock that *loses* 5% over a week isn’t twice as bad as one that takes two weeks to do that: I still lose the 5% either way and spending more time on the loser means lost opportunity for gain on some new purchase.  So maybe %/day is like the P/E ratio in that it is meaningful only when positive.

% gain, less 0.04%/dayRealized
Symb Buy Fri Mon Tue Wed Thu Fri gain
 
DRETF DE16 -0.5 0.0 1.1 1.6 3.5 4.7
ETN FE03 3.2 2.5 0.8 2.1 1.6 1.6 +1.3%
GIII FE03 0.4 -3.0 -3.0 -1.8 -1.2 -1.0 −1.2%
ATML FE04 7.4 8.6 5.8 5.0 4.9 4.9 +4.8%
MKL FE04 1.1 0.5 0.2 0.4 0.7 0.9 +0.4%
CAT FE07 3.4 3.1 2.4 2.6 2.5 3.7
SPY FE08 0.3 0.5 0.2 0.8 0.6 0.6
TNA FE09 2.5 4.0 1.9 3.0 4.1 4.2
ADTN FE14 -1.3 -2.5 -2.6 -0.5 -1.4 −3.7%
ENTG FE14 7.2 3.7 3.6 3.8 3.2 +4.1%
LINTA FE17 0.8 0.6
CR FE18 0.1
MMS FE18 1.9
ARE FE18 -0.9
(Average) +1.2%

The second table shows an alternate view of the data.  These are total gains since purchase, minus 0.04% for each day (including hold days and settlement days) so a score of "0.0" means the stock is on target to yield 10%/year.  If the price doesn’t change then this table will show a declining score because of the subtraction; if the price goes down then the score will go down even more.

Some of the numbers here don’t match up with the previous table.  For example, DRETF now shows higher historical closing prices than I actually saw on those days; maybe this is an exchange-rate issue?

As usual, grey indicates a hypothetical post-sale score.  A green half-bar means I sold half the shares; at that point the price of those shares is frozen while the other half continues to track the market price (there is also an increase in trading cost, which reduces gain slightly).  The green ¾ bar for ENTG means I sold another quarter of the shares, leaving only ¼ still tracking the market price.

The average is weighted by the sizes of the bets (GIII and MKL were half as big as the others).  I had a nice week, as far as realized gains are concerned!  The problem was unrealized losses: as shown in this table, many stocks held from last week were worth a lot less on Tuesday than they were last Friday.

The new “2-for-1 strategy” (sell ½ the shares if price rises by ½ the initial stop depth) makes this pie chart of my current holdings look very busy!  The lowercase stock symbols have been sold and will rejoin cash on the date indicated.  I’ve decided that it doesn’t make much sense to use the 2-for-1 strategy a third time, leaving only ⅛ shares still held, so if the stock goes up by 150% of initial stop depth then I’ll sell the remaining quarter of the initial purchase.

Notes on individual stocks:

GIII: My sell-stop was triggered by a $36.60 bid that no one took, causing me to get the lowest official price of the day which was $37.02.  This wasn’t so bad because I would have gotten even less if the stop were triggered the following day by an actual trade.

MKL: My $414.14 sell-stop was triggered by a bid no one took, causing my shares to get sold for $416.66 which was *below* the official low price of the day!  There is something fishy here.  Why didn’t my sale count for the day’s low?  Possibly because I sold very few shares of this very expensive stock.  After shaking me off, MKL proceeded to rise 0.7% that day.

ADTN: Gapped down at the open on Thursday, shaking me off, then proceeded to rise 2% that day!  On Wednesday I had thought up a new rule: if a stock falls to 50% of the initial stop depth, then sell half of it.  But ADTN’s closing price on Wednesday was a few pennies more than that cutoff so I didn’t sell.  As it turned out, the result would have been the same either way.

ATML: I would have made even more had I implemented the 2-for-1 rule earlier!

ENTG: I was unable to set Thursday’s pSAR value as the stop for Friday because Thursday's closing bid was lower than that.  In such cases, I can either keep the previous stop or sell at market price.  In this case, I decided to sell.  It went down after my sale!  Win!

TNA: It has risen enough for another partial sale, but it was a half-sized bet to begin with so I have put in a market order to sell all remaining shares.  This suggests that I will need a new rule: when should I buy TNA again to start another round?  Ditto for ATML, which has frequent run-ups and will likely have another one soon.

CAT and MMS: Both of these have risen enough for another sale, so I have put in sell orders for ¼ CAT and ½ MMS.

* * * * *
I called Schwab to ask for “active trader” status, because that will allow me to use sell-stops that trigger on actual TRADE prices, not on hypothetical BID prices.  This should become effective on Tuesday.  I might also try using the “bracket” type of order to sell-stop if the price falls or sell-limit if the price rises.  But the demo of the active-trader system makes it look like bracket stops always use those stupid BID prices, which bit me twice this week.

Tuesday 15 February 2011

Older stuff

For older entries, see my furry blog: http://pyesetz.livejournal.com/tag/finance.