Week of 2012 JA 13

SPY rose by 0.9% this week, while my account fell by 0.1%. 

The market tends to make a spike-low every 35 to 40 days.  There have been 33 trading days since the last spike-low on Nov. 25th.

		│	Daily % gain size																	│	Max % loss			│	Results
Symbol	since	│	Fri			Mon			Tue			Wed			Thu			Fri		│	Beg		End	│
RWM	JA06	│	+0.1	6		-0.5	6		-1.3	10		-1.8	10		-2.0	10		-1.5	10	│	−7.8		−6.9	│
SPY		│	+1.8			+2.0			+2.9			+2.9			+3.2			+2.7		│				│
me		│	+0.0			-0.0			-0.1			-0.2			-0.2			-0.1		│				│

For this year, the SPY and me rows in the table now start at +0.0 on JA 01, rather than restarting at +0.0 each week.

RWM: Gapped down on Tuesday, so I bought more at a lower price (which effectively tightened my stop-loss).  I had a limit-order to buy a third tranche if SPY rose to $130.50 (just below major resistance), but that didn’t happen.
      Next week: consider tightening the stop to 5%.  The 8% stop is for index funds to be held for several months, but I have doubts that I can really bear to hold this thing so long.  If I’m going to sell it at the January spike-low, the correct trailing-stop value is 5%.

MACD-based swing trades

Still no swing trades this week!  Hopefully the Minor [3] wave has just begun today and will trend downward next week.  PPO-signal is now only 0.18 and PPO-histogram is -0.20, so my robot-trader won’t buy anything unless there is a sharp drop of maybe 2% to trigger the delta rule.

Week of 2012 JA 06

SPY rose by 1.8% this week, while my account did not change.  The chart has been rescaled and rebased so "0% change" now refers to the values at the beginning of this year.

		│	Daily % gain size																	│	Max % loss			│	Results
Symbol	since	│	Fri			Mon			Tue			Wed			Thu			Fri		│	Beg		End	│
IAU	DE16	│	-2.0	10					+0.6	10		+1.1	0							│				│	−4.1%
RWM	JA06	│																+0.1	6	│	−8.1		−7.8	│
SPY		│	0.0						+1.6			+1.8			+2.0			+1.8		│				│
me		│	+0.0						+0.0			+0.0			+0.0			+0.0		│				│

Markets were closed on Monday, in observance of New Year’s Weekend.

RWM: Bought because DanEric has been saying since DE 23 that the market would turn South on or about JA 09, which is next week.  He seems to have been on a roll these last three months: his “Minor [2] double-zigzag with ending diagonal” analysis has been quite accurate all along.  I placed a fairly small bet, mainly because PPO-signal is not very high right now so my automated system won’t go short until the market drops significantly.

MACD-based swing trades

Still no swing trades this week!  Maybe next week, as Minor [3] begins.

Week of 2011 DE 30

SPY fell by 0.7% this week, while my account fell only 0.4%.  For the year, SPY fell less than 1% while my account fell more than 17%.  I should’ve just put my money in SPY and left it there all year!

I have rebased the chart so the beginning-of-2011 values are now 0% instead of 100%.  Next week I will rebase the chart again so this week’s ending values (which are also start-of-2012 values) will correspond to “0% change” on the chart.

		│	Daily % gain size																	│	Max % loss			│	Results
Symbol	since	│	Fri			Mon			Tue			Wed			Thu			Fri		│	Beg		End	│
IAU	DE16	│	+0.8	10					-0.1	10		-2.6	10		-3.1	10		-2.0	10	│	−4.0		−4.0	│	−4.1%
SPY		│	0.0						+0.1			-1.2			-0.2			-0.7		│				│
me		│	0.0						-0.1			-0.3			-0.4			-0.4		│				│

Markets were closed on Monday, in observance of Christmas.

IAU: Stopped out on Thursday, making a lower low.  This is not a good time to be holding gold.  I will wait awhile.

Going into New Year’s, my holdings are now 100% cash (including unsettled cash from sale of IAU).  Next year my “gains this year” and “gains from transactions initiated this year” will be the same, unlike this year when there was a 2% unrealized gain carryover from 2010 that kept tripping me up.

MACD-based swing trades

My automated trading system for TNA/TZA swing trades is live and seemingly operational!  But there were no swing trade set-ups this week, so it hasn’t actually done anything yet.

New rule: Do not buy TNA when PPO-signal > -0.7; do not buy TZA when PPO-signal < +0.55.  Along with the “tweaks” below, for TNA this rule improves the 2011 win:loss ratio from 22:14 to 16:7, avoiding many losses while keeping most gains.  Overall profit increases from 2.52% to 2.53%, so basically no change.  For TZA this improves the win:loss ratio from 11:8 to 9:3 and increases profit from 7.25% to 7.28%.

Tweaks: Increase TNA delta from 1.1 to 1.2; increase neg-delta from 0.35 to 0.40.  Decrease TZA neg-delta from 1.45 to 1.15.  In part these tweaks are needed to reduce the negative impact of the new PPO-signal rule, because otherwise the delta rule would just buy the stocks anyway!

The TNA swing-trading rules seem to work well now.  Average gain per swing is 4.6% for the last 12 months, with roughly 3:1 odds. For the last 18 months (all the data I have), the average gain is also 4.6% but the odds are only 2:1.

The TZA rule needs more work.  Average gain is 18% for 12 months, with 3:1 odds, but for 18 months the average gain is only 10.4% and again the odds are 2:1.  There are still several nasty -5% losses, such as the -5.6% loss for October 27th.  For TNA, I was able to eliminate the nasty -3% loss in October via the above tweaks; maybe the same can be done for TZA with some more tweaking.

Week of 2011 DE 23

SPY rose 3.8% this week, while my account fell 0.1%.  Next week is expected to be about the same.

		│	Daily % gain size																	│	Max % loss			│	Results
Symbol	since	│	Fri			Mon			Tue			Wed			Thu			Fri		│	Beg		End	│
TZA#0a	NO30	│	-2.5	9		+2.6	9		-10.0	9		-11.2	0							│				│	−4.9%
TZA#0b	NO30	│	+2.5	9		+7.9	9													│				│	+9.7%
TNA	DE15	│	+1.6	10		-3.8	10		+8.1	10		+9.0	10		+11.5	0				│	−10.0		−10.0	│	+0.0%
IAU	DE16	│	+0.2	10		-0.0	10		+1.3	10		+1.3	10		+0.6	10		+0.8	10	│	−5.4		−4.0	│
SPY		│	0.0			-1.0			+1.9			+2.1			+2.9			+3.8		│				│
me		│	0.0			-0.2			-0.0			-0.0			-0.1			-0.1		│				│

TNA: Sold due to drop in MACD, but then the market “melted up” for the Santa Clause rally.  MACD does not do well with melt-ups, but TRIX does.  Perhaps I should use both indicators?

IAU: Doing okay.  Sell next Friday unless it does very well and my stop rises above break-even.

Automated trading: I have written up a system that accesses Schwab’s website using a web browser and some Greasemonkey scripts.  Next week I’ll let it run “dry” to shake out the bugs, then maybe give it some money the following week.  (Might have to wait an extra week if the Santa rally continues and there are no MACD swing trades for the rest of the year.)

MACD-based swing trades

Ticker Symbol		Buy date			Buy price			Sell date			Sell price			Profit
		Model	Actual		M	A		Model	Actual		M	A		M	A
TNA		DE 15 14:00	DE 15 16:50		$40.90	$41.00		DE 19 12:00	DE 19 12:10		$40.75	$41.10		-0.6%	+0.0%
TNA#2		DE 20 11:00	skipped		$43.62			DE 21 13:00	skipped		$43.27			-1.0%	+0.0%

TNA: As usual, the model bought for a better price than I did, but I sold for a better price.  Managed to break even, during a week when this ticker rose a lot.

TNA^#2: Skipped this trade bacause PPO-signal was positive.  Avoided a loss.

Week of 2011 DE 15

SPY fell 3.5% this week, while my account rose 1.8%.

		│	Daily % gain size																	│	Max % loss			│	Results
Symbol	since	│	Fri			Mon			Tue			Wed			Thu			Fri		│	Beg		End	│
TZA#2a	NO30	│	-10.7	8		-6.9	8		-1.3	9		+2.8	9		-0.2	9		-2.5	9	│	−13.6		−11.2	│	−4.9%
TZA#2b	NO30	│	-6.0	9		-2.1	9		+3.8	10		+8.1	10		+4.9	10		+2.5	9	│	−9.2		−6.6	│	+9.7%
TNA	DE15	│													-0.8	10		+1.6	10	│	−15.2		−10.0	│
IAU	DE16	│																+0.2	10	│	−5.7		−5.4	│
SPY		│	0.0			-1.5			-2.4			-3.4			-3.1			-3.5		│				│
me		│	0.0			+0.6			+1.5			+2.2			+1.9			+1.8		│				│

IAU: Bought because ZeroHedge recommends it.  They say to hold it for about two weeks.  Toby Connor also recommends gold (hold until it goes stratospheric), but he’s been wrong a lot recently, so I’ll stick to my usual 5½% trailing stop.  If I get only a 38% retrace of the recent drop, that would be a 2.5% profit.

Several times now I have missed trades because I have other things to do in life besides check the market every hour on the hour.  If I can manage to make some money next month, perhaps I should look into a way to automate the trades coming from my model, since it seems to be doing a better job than I am!  Unfortunately, Schwab doesn’t seem to support automated trading.  I found a news article suggesting that they sold their automatic-trading system to UBS.  It is difficult for me to leave Schwab because I have an IRA retirement account with them and am no longer a US resident.

MACD-based swing trades

Ticker Symbol		Buy date			Buy price			Sell date			Sell price			Profit
		Model	Actual		M	A		Model	Actual		M	A		M	A
TZA#2a		DE 07 11:00	NO 30 09:30		$28.31	$29.97		DE 16 11:00	DE 16 11:10		$28.39	$28.58		+0.0%	‒4.9%
TNA#0		DE 09 14:00	skipped		$45.62			DE 12 12:00	skipped		$42.71			‒6.6%	+0.0%
TNA		DE 15 14:00	DE 15 16:50		$40.90	$41.00		not yet	not yet

TZA^#2a: Bought way too early; sold on time.

TNA^#0: Skipped this trade and avoided a significant loss.  Maybe my model should avoid buying TNA when PPO-signal is positive.  That wouldn’t help overall returns much, but it would avoid a bunch of loser trades.  I don’t need any more aggravation!

TNA: Normal swing trade in progress.  Bought in the aftermarket because I was in a business meeting at 2pm, but ended up getting roughly the same price anyway.

Surfing the oscillations

Last week I wrote: “The sell-limit order for a 38% retrace on that second tranche still hasn’t triggered. Now it’s a race to see whether the tranche sells before I get a ‘buy more’ signal from the model or get stopped out.”  Well, the race was won by the sell-limit order!

So I split the tranches of TZA^#2 back into separate trades: the purchase of TZA^#2a was eventually approved by my MACD model and so I sold it per model, while TZA^#2b never did get approval and was sold at a Fibonacci retracement as shown on this chart.

Week of 2011 DE 09

SPY rose 0.9% this week and is now slightly above its value at the start of the year, while my account fell another 0.8% to its lowest value ever.

		│	Daily % gain size																	│	Max % loss			│	Results
Symbol	since	│	Fri			Mon			Tue			Wed			Thu			Fri		│	Beg		End	│
TZA#2	NO30	│	-2.8	18		-7.8	17		-8.2	17		-7.8	17		+1.0	17		-8.2	17	│	−11.3		−11.3	│
SPY		│	0.0			+1.1			+1.1			+1.5			-0.7			+0.8		│				│
me		│	0.0			-0.8			-0.8			-0.7			+0.6			-0.8		│				│

Euro news: The Euro conference has failed.  The plan for tighter union put forward by Germany+France was officially rejected by the UK and unofficially by Finland, the Netherlands, and perhaps others.  If you believe that ”the news moves the markets”, you would have expected prices to go down today on this really bad news.  If you believe in wave theory, today’s huge run-up was predictable.  Unfortunately, the prediction calls for prices to go so high next week that my TZA shares will stop out, then prices will drop sharply.  I’m still hoping that my shares will survive because the price spike will be weak (maybe today was the top?).

MACD-based swing trades

Ticker Symbol		Buy date			Buy price			Sell date			Sell price			Profit
		Model	Actual		M	A		Model	Actual		M	A		M	A
TZA#2		DE 07 11:00	NO 30 09:30		$28.31	$29.97		not yet	not yet
		not yet	NO 30 15:50			$28.50
TNA		DE 09 14:00	not yet		$46.62			not yet

TZA^#2: Yay!  I got a “buy” signal before getting stopped out!  But my attempt to beat my own model has failed pitifully.  Should have waited.  The sell-limit order for a 38% retrace on that second tranche still hasn’t triggered.  Now it’s a race to see whether the tranche sells before I get a “buy more” signal from the model or get stopped out.

TNA: Didn’t buy because I was out driving around town today, so trading was inconvenient.  Maybe buy it Monday if prices don’t come down a lot.  Maybe buy a double-tranche to neutralize the TZA and prevent my account balance from falling any further.

John Murphy suggests that the MACD signal should be negative when the histogram is positive in order to have a proper “buy” signal.  I definitely found this to be the case for my attempts at day-trading a few weeks ago.  And the MACD signal is currently positive so this would suggest not buying TNA.  However, at the hourly level the MACD-signal doesn’t seem effective.  Let's suppose that the tranches consist of 10% of my account put into SCHA (which is unleveraged and has no trading costs, so the math is easy).  Over the last year, my model would have produced 22 profits and 12 losses for a total gain of 3.5%.  If trades with negative PPO-signals were excluded, there would have been 17 profits and 8 losses for a total of 3.3%.  Not useful!

Week of 2011 DE 02

SPY rose 6.8% this week, almost returning to its value at the start of the year, while my account fell 2.9%, to its lowest value ever.

		│	Daily % gain size																	│	Max % loss			│	Results
Symbol	since	│	Fri			Mon			Tue			Wed			Thu			Fri		│	Beg		End	│
TZA#1	NO18	│	13.9	23		-2.1	20		-1.5	20		-19.0	20		-17.4	0				│	−3.1		−3.1	│	−0.6%
TZA#2	NO30	│										-3.6	18		-1.6	18		-2.8	18	│	−15.2		−11.3	│
SPY		│	0.0			2.7			3.0			6.9			6.9			6.8		│				│
me		│	0.0			-2.5			-2.5			-3.0			-2.7			-2.9		│				│

World news: On Monday, the central banks of the USA, the UK, the Eurozone, Switzerland, Canada, and Australia decided that they would accept each other’s currency as collateral for swaps, so if you are a bank and all you have is Euros and nobody trusts you enough to loan you US$, the US government will be your lender-of-last-resort.  The stock market surged!  Just like it did the previous times this was tried, in 2008 and 2010.  These intervention surges generally last about a week, so it should be just about over by now.
      An interesting side-note: the decision was made on Monday, but not publically announced until Wednesday, so the crony capitalists of the world could front-run the news at the expense of everyone else.  Increasingly these days, The Powers That Be are not even trying anymore to hide their corruption.
      A broad 7% rise in the market, based solely on an artificial intervention that is only good news for banks and not for industrials, suggests that only the banks matter now; everyone else has stopped trading.

MACD-based swing trades

Ticker Symbol		Buy date			Buy price			Sell date			Sell price			Profit
		Model	Actual		M	A		Model	Actual		M	A		M	A
TZA#1		NO 17 13:00	NO 18 10:42		$31.56	$31.73		NO 28 11:00	NO 28 13:15		$34.30	$34.61		+3.6%	−0.6%
		NO 21 11:00	NO 21 11:43		$34.64	$34.71
		do not buy	NO 25 11:22			$37.80
TNA		NO 28 11:00	skipped		$38.21			NO 29 15:00	skipped		$38.01			−1.0%	+0.0%
TZA#2		do not buy	NO 30 09:30			$29.97			not yet
		not yet	NO 30 15:50			$28.50

TZA^#1: Shouldn’t have bought that third tranche.  Too greedy!  Would have been profitable otherwise.

TNA: Skipped this post-intervention swing.  Avoided a loss!

TZA^#2: This really isn’t a swing-trade, since I bought it when PPO was giving a strong “sell” signal.  But it might become a swing trade next week!  As of 4pm Friday, IWM's PPO = -0.56, which would be a “buy” except that my model ignores the 4pm and 10am values.

Surfing the oscillations

Elliott Wave Theory usually consists of counting squiggles and grouping them into threes and fives.  One notable exception: the 3-in-1 rule.  The week starting OC 24 was a “setup week”, followed by three ”inside weeks” and then the week starting NO 21 was a “trigger week” showing a downside breakout.  We should normally expect that that prices will continue falling for months.  But that assumes no intervention by the central banks!

On Monday I assumed that the bull-run was surely almost tapped out, because the lower orange bar was so close.  So I put in a limit order to buy TZA if prices got up to roughly $71.50 on this chart.  On Wednesday I found myself the proud owner of TZA shares, bought for much better than my limit price!  Now what?  I used the blue bar to guess what the new can’t-go-any-higher price would be and put in another limit order, which triggered just before Wednesday’s close.  On Thursday and Friday the market continued even higher.

DanEric thinks the market is going down next week, but will just be a short correction and then the Christmas rally will start in earnest.  I think the market will spend the rest of the month oscillating between the NO 28 lows and the DE 02 highs, like it did back in August.  I suppose this will depend on what the Euro leaders come up with at next Friday’s save-the-Eurozone meeting.  Just in case, I have a limit order to sell half the TZA if prices retrace 38% of their recent ramp, which would again be about $71.55 on this chart.  DanEric thinks this could happen as soon as Monday.