10:15 am EDT:  Market liquidity, as measured by proxy through the ES via a proprietary measure, had improved over the last few weeks, but just deteriorated significantly.  The last time it approached these levels from the downside (greater to less liquidity) was May 6.  This doesn’t mean there will necessarily be another flash crash, but that it is very easy to push prices around on low volume.  Caution is warranted with all positions.

The best time to be long continues to be the morning, while being short in the afternoon has paid off recently, especially during the 1:30 pm to 3:00 pm EDT period. 

In the morning report, we said equities have been sideways to down in the last week of the month since August 2009.  More precisely, it is the week that coincides with the long term Treasury auctions in the 2, 5 and 7 year tenors that tends to exhibit this behavior.  Usually, this is the last week, but not always.  The first of the auctions is the 2 year, which usually occurs on a Tuesday, but sometimes a Monday due to holidays.  Below highlights 2 year auction dates in magenta.  In addition, FOMC Announcement dates are highlighted in cyan.  With an announcement scheduled for next Wednesday, it is also instructive to look at the two occurances (in September 2009 and January 2010) when there was an announcement in an auction week.  This gives next week a bearish seasonality until the last of the auctions, the 7 year, is over on Thursday.  GDP on Friday could be a catalyst for a continuation of the rally, however. 

As always, seasonalities should be taken with a grain of salt as any number of other factors can override.

Accelerated/Decelerated Midas (ADM) Fan ™ / Accelerated/Decelerated VWAP (ADV) Fan ™

Most have at least heard of VWAP, or volume weighted average price.  It is a measure of the average price paid over a period of time, and can be implemented a number of ways.  Often, the calculation is reset daily, though in the MIDAS Method that was created by Paul Levine it is launched from important reference points in time and allowed to continue (called a Midas curve). 

The original formula is simple.  For each price bar, the following is applied:

PV = PV + (Price * Volume)
 CumeVolume = CumeVolume + Volume
 VWAP = PV / CumeVolume

where Priceis usually (High + Low + Close) / 3.  Coles and Hawkins of MIDAS Market Analysis have also discovered it is permissible to use just the High in a downtrend, and the Low in an uptrend, which will cause the Midas curve to follow price a bit more closely.  They will reveal additional insights and modifications in their upcoming book. 

For now, we have modified the formula to accelerate or decelerate the curve above/below the main Midas curve by incrementally increasing/decreasing each new bar’s contribution as follows:

PV = PV + (Price * Volume * Factor * Count)
CumeVolume = CumeVolume + Volume
VWAP = PV / CumeVolume
Count = Count + 1

By including a range of values for Factor, such as 1.00005, 1.000010, 1.000015, …, 1.000100, we can establish an entire Accelerated/Decelerated Midas support and resistance fan (ADM Fan).  One could think of it loosely as a volume-based Gann fan.  While this works with time charts, it has empirically been found to work best with volume charts, wherein each bar represents a fixed number of shares/contracts.  A good rule of thumb for the Factor range is that the minimum should be within an order of magnitude of 1 / ContractsPerBar, which depends on how the chart is set up.  The maximum can be three orders of magnitude (x 1000) higher and will depend on the steepness of the trend.  It is usually not necessary to curve-fit the increment and range as long as the launch point is correctly chosen.

Above shows the ES with an ADM Fan launched from 11:56 am EDT on February 25, 2010.  We chose this time, and not the time corresponding with the actual low, because it was from here that the market launched higher.  The blue line is the original Midas/VWAP curve, which captured price early.  As price displaced from the Midas curve, Accelerated Midas curves took over and became both support and resistance.  None of the Decelerated Midas curves was reached.  However, for a good example of what this looks like, we will show the ADM Fan launched from the actual aforementioned low at 9:45 am the same day.

Above, the third Decelerated Midas curve captured the 11:56 am retest of the low, which was the launching point for the ADM Fan in the first chart.  This Fan also captured highs and lows fairly well. 

Because the ADM formula weights bars increasingly more as time progresses, the fan, as its name implies, spreads out.  By decreasing the increment amount by which the Factor is increased, we can generate more support and resistance lines that become relevant over time.

Below shows an even lower Factor increment amount that produces yet more levels, along with a new ADM Fan launched on the overnight low of March 4, 2010.  When confluence occurs in curves from different launch times, support/resistance is more powerful.

It would be easy for a chart to get cluttered (if it is not already), so selection of launch points is of great importance and an area of current development.  In general, important highs and lows work well, but sometimes it is best to launch, not from the actual low, but the point at which price takes off (as demonstrated earlier).  Launch points from areas that are important support and resistance levels based on other methods also work well.

Below is another example, where an ADM Fan is launched in the EuroFX futures contract near the 4:20 am low on March 2, 2010, which was a new yearly low. 

Below shows a reduced Factor increment that in turn produces a greater number of relevant support and resistance levels.

Reverse Midas ™ / Reverse VWAP ™

The indicator at the bottom of the above picture is Reverse Midas/VWAP, and is calculated as its name implies.  For each given bar, iterate backwards, calculating VWAP until zero is reached.  Buyers or sellers at this bar were at break even when the subsequent bar from which the backwards iteration began was reached.  Once the first equilibrium/breakeven point is reached, we continue to iterate backwards until the next is encountered, and repeat the procedure for a total of 5 (arbitrarily chosen) times.  The volume levels between the points are then plotted below price.  Also important (not shown), is the maximum average profit and loss achieved over these intervals, which is akin to the Active Boundaries ™ method described by Pascal Willain in his book Value in Time. 

Often, as a trend progresses, equilibrium points will build and be revealed as clusters by the indicator, which will eventually tend to stop the move.  When the clusters disappear, there are fewer traders fighting price and a new trend can emerge.  A similar phenomenon occurs when there is a cluster reached of extreme average profit/loss.  Again, this is not shown, but will be demonstrated in future posts as we further elaborate on these concepts.  Also to be explored is the importance of certain volume levels that emerge repeatedly over time from these calculations, which can be relevant to other parts of the Midas Method, such as with Top/Bottom Finder curves.

Well, it took a bit longer than we thought, but the US Dollar Index has broken major support.  Two downside targets are highlighted below that each contain long term pivot and Fibonacci support.  With the ES close to entering the 1127 to 1147 range from January, any further downside in the Dollar would help equities power through resistance. 

Click for larger image.

Below updates a previous chart we posted that breaks up the day into time segments (all times EDT) and plots net points for each period.  Since the January 2010 highs, the first hour is where most of the losses have occurred, even during the most recent upswing.  Since the February low, each other period in the rest of the trading day has been net profitable, though the closing hour the least so.  The 1:30 to 3:00 pm time of day has been the most profitable intraday period for longs since the 2009 rally began.

Looking farther back to 2001, we can see glean some interesting information.  The first hour of trading was consistently profitable beginning late 2003 during the last bull market.  It posted its high in July 2007, three months before equities actually topped.  It also bottomed concurrently with the markets in March 2009.  Similar to 2003, it was unable to trend up for most of 2010, but took off late in the year.  As previously noted, it has since retreated and we believe will need to start turning profitable if the January highs are to be taken out.

The only other time period that was consistently profitable during the 2003 to 2007 bull run was the overnight gap.  If it were to break its current trend line, that would be a very bearish sign.  What also emerges is that, contrary to popular believe, a profitable closing hour is not necessary to sustain a bull market, but may be necessary to start one.  The closing hour’s high was put in January 2004 and trended down thereafter.  The closing hour’s high in the current 2009 rally was established in September.  It may well be that smart money now trades at the open.

Accordingly, if we start seeing overnight gap-ups and profitable first hours, things may have turned around for equities.  If not, we’ll probably see lower prices in March.

On the heels of the surprise discount window rate hike late last week, and on the eve of Bernanke’s Congressional testimony, speculation abounds as to the when and where of the next round of tightening.  We need look no further than the US Treasury press room, as it has announced today a revival of sorts for its Supplementary Financing Program (SFP).

Remember the SFP?  It’s back, though it really never went away.  Originally created in September, 2008 to provide a pool of funds that could be drawn upon by the Fed in emergencies without adding to excess reserves (before the Fed had the power to pay interest on excess reserves), the SFP hit its peak amount in November, 2008 at $558.9 billion.  Thereafter, it was quickly drawn down to about $200 billion by February 2009, where it remained until Treasury ran into debt ceiling issues in September and announced it would be wound down to $15 billion.  In fact, by January 6, 2010, only $5 billion remained. 

Today, Treasury announced as follows:

February 23, 2010
TG-560

Treasury Issues Debt Management Guidance on the
Supplementary Financing Program

WASHINGTON –The U.S. Department of Treasury today issued the following statement on the Supplementary Financing Program (SFP):

“Treasury anticipates that the balance in the Treasury’s Supplementary Financing Account will increase from its current level of $5 billion to $200 billion.  This will restore the SFP back to the level maintained between February and September 2009. 

This action will be completed over the next two months in the form of eight $25 billion, 56-day SFP bills.  Starting tomorrow, SFP auctions will be held each Wednesday at 11:30 a.m. EST, unless otherwise noted.”  

###

We speculated after the September 2009 wind down announcement that it (1) would provide another $185 in liquidity for risk markets as the cash management bills that financed the program were not rolled over and returned to primary dealers, and (2) would increase demand for short term bills.  Since the process will now be reversed, it is reasonable now to believe the outcomes will be reversed as well.  Indeed, as Zero Hedge noted in a similar story earlier, demand is already disappearing from indirects in short term bill auctions. 

With the brunt of the $200 billion cash management bill sales expected to be picked up primary dealers, this will have the same effect as adding up to $200 billion to bank nonborrowed excess reserves (NBER) on deposit with the Fed.  As bank NBER is just north of $1 trillion, a 20% increase over eight weeks in the amount of non-borrowed money locked up at the Fed is material.  At a time when Agency and Agency MBS are drawing to a close, and with M2 money supply flat, this de facto tightening move is a bit alarming. 

Further, using the 13 week T-Bill rate of 0.1% as a proxy for the shorter duration 56 day (8 week) bill, it yields less than half the 0.25% paid by the Fed on excess reserves.  Accordingly, even if existing excess reserves are used to finance the SFP, resulting in a net wash in money locked up at the Fed, the marginal profit provided by this carry trade and so needed by the large banks will be materially diminished.  Under the same net wash scenario, this move could also be a precurser (test run?) for the term deposit facility proposed by the Fed.

For anyone who doubts the intent of these actions, we need only revisit the original press release by Treasury:

September 17, 2008 
 
Today, the Treasury Department announced the initiation of a temporary Supplementary Financing Program. The program will consist of a series of Treasury bill auctions, separate from Treasury’s current borrowing program, with the proceeds from these auctions to be maintained in an account at the Federal Reserve Bank of New York. Funds in this account serve to drain reserves from the banking system, and will therefore offset the reserve impact of recent Federal Reserve lending and liquidity initiatives.

As the Fed now has myriad tools to offset the reserve impact of liquidity initiatives and is unlikely to restart such initiatives in the near term, this is purely and simply a reserve draining mechanism that will at best erode bank profits and, at worst, shrink an already precariously perched money supply.  We will analyze Fed statistics over the coming weeks and update as to which is the more likely scenario.

It’s important not to become too bearish in the short term on long term news, especially on a net down day in equities.  For those that subscribe to our daily reports, this does not affect our view that the US Dollar is topping this week and due for a modest 38% to 50% correction of the recent up leg.   A concurrent equities rally would still accompany, but we are now less confident in its ultimate potential.  

Only this morning we wrote:

…[W]e have a quiet news week and the markets are sensitive to unscheduled news.  Overnight, equities are up on some bullish rumors regarding a Greek bailout.  Some have compared the situation to the Dubai events in late November that led to a quick selloff and rebound in world equities.  However, this situation is not to be taken lightly, as a Greek default would be three times as large as the Lehman bankruptcy, and could quickly devolve into another global crisis of confidence.  Accordingly, the markets are at the whim of the ECB.  If it bails out Greece, there will probably be a large short covering rally.  If it lets Greece default, there will probably be another selloff.  And, if it does nothing and Greece muddles through for the time being, there will probably be a series of minor rallies that lead to larger selloffs.  The moves generated by the first two scenarios will be very swift, so swing traders will need to be prepared to react just as quickly…

When news broke at about 11:30 am EDT that there was an agreement in principal for a bailout, equities rallied and the US Dollar fell, as expected.  However, a mere hour later, after the ES had rallied 19 points to 1077, Germany countered by saying it was not a done deal and there would be significant strings attached.  Accordingly, there is still much uncertainty in the markets.  For now, what likely would have devolved into a return to the 1040′s has been averted.  In the end, we believe a bailout with nominal strings attached (to save face) is very likely, but the intervening journey in the markets will be volatile as the details are filled in over the coming days and weeks. 

The 1080 to 1083 is the first critical resistance level that swing shorts will need to defend.  There is a historical tendency to clear important resistance levels overnight, evidenced by the fact that the gap accounted for fully 32% of the rally in the S&P 500 that began in March 2009.  Combined with the current news being generated overseas, if this area is to be exceeded, US traders should be prepared to wake up to a market that has already cleared it rather than experience it intraday.

There will be plenty of opportunities during trading hours, however, and a savvy daytrader can capitalize on these movements by correctly reading market signals, regardless of knowing the actual news that’s driving the markets.  A simple five minute candle chart with volume of the ES warned that the rumor would lead to a sustained rally when it closed nearly at its highs (within a tick) on high volume.   Ideally, volume would have been at least 100,000 (actual about 91,000), but the 8 point range that convincingly broke the downward trendline was sufficient to generate follow through short covering.  Also important was that the ES looked like it was headed for trouble and sentiment was very negative following the failure just above the previous day’s high.  The entry can be made on a stop basis one tick beyond the big range bar, with a two point stop loss. 

In general, the larger the wick or shadow of a big range candle, the less likely a continuation of the move is.  This so-called indecision area is just that–it conveys doubt and will encourage profit taking and counter trend traders, which will tend to halt the move.  This is why a short based on the big-down 12:45 pm candle was not a good candidate for a continuation move (besides the fact that the stop sell entry signal was not triggered).  The two point shadow at the bottom was enough to make a material retracement to the 50%-61.8% fib box likely.

The flipside to potential entries is that, in this environment, exit stops placed on day trades are crucial because it is easy to get caught on the wrong side when surprise news is announced.  Indeed, a long entered on the above basis would have given up 100% profits on the 12:45 pm bar.  Accordingly, on steep moves, a simple trendline break can be a profit taking cue, as can a move that exceeds an interim pivot bar on the 5 minute chart.

Moves like today do not occur frequently, but when they do, often follow a predictable pattern.  We should see more in the coming weeks, so be prepared.

For a daily battle plan and intraday updates, register free at our site.

The 114.18 high volume level has now been reached.  In the chart below it does not appear to be a high volume level because the area between 111.15 to 114.18 has recently been filled in.  The next target is 116.17; however, a drop back to 111.15 is possible if SPY stalls here.

Click for larger image.

We have compiled the following long term charts that show points accumulated overnight (just ahead of) the Employment Situation report, then during the day session on the day of the report, and the combination of the two. 

 
Above is Feb 94 to Dec 09, and below is a zoom in of Sep 08 to Dec 09.

For 2009, Jul 09, Oct 09 and Nov 09 were negative gap days.  The day sessions were positive Aug 09 to Nov 09, but negative Dec 09.  What’s interesting in the longer term chart is that day sessions were positive in SPY throughout must of the late nineties bull run, but were not so during the subsequent bull run.  Perhaps that’s indicative of the jobless recovery.  SPY bottomed in Oct 02 on a higher low in the day session.  After that, there was a rally in day sessions, but then there was a resumption in the down trend until Oct 08, that lasted even during the bull market into Oct 07. 
 
Meanwhile, the gap stayed mostly flat and resumed its uptrend in early 2005.  Further, during the entire recent down leg, the worst drawdown in the gap was 2.34 SPY points, with most of the losses to SPY overall coming during the day session.  Accordingly, holding overnight ahead of the report (at least since mid-2003) appears to be more profitable than holding during the day.  It should be noted that there are outliers where SPY has gapped down ~1.50 points and it is not uncommon to have ~1.0 point gap downs, so caution should always be exercised when holding into the report.