Let's take this Friday afternoon to add to the knowledge base:

---------

Price Discounts Everything: How To Reconcile A Data-Driven Price Movement Against Technical Analysis

Over 100 years ago, a principle was suggested by Charled Dow which eventually became central to technical analysis.  In fact, several principles suggested by Dow that have more than withstood the test of time in  multiple aspects of the financial markets, thus coming to be known as "Dow Theory."  Today's discussion focuses on the following principle: "Price Discounts Everything."  In other words, current market prices reflect the digestion of past information, the reaction to current information, and the expectation of future information.  But it's not as simple as saying that the expectation of a particular economic indicator is baked-in to current trading prices.  On a more esoteric level, it's not just the expectation that the indicator will meet it's consensus, but also reflected in the price is the "unknown" in the sense that it might vary greatly from expectation. 

It's from that point that more specialized branch of technical analysis diverges from Dow Theory.  If the markets were truly pricing in the actual future information, prices would have no reason to move.  But because that information and therefore prices, have variability, technical analysis seeks to apply studies to that variability in the hopes of gaining further understanding and potentially even another modicum of certainty regarding the expectation of the future. 

For instance, if today's expectation of home sales was far lower than the actual print, prices cannot have already reflected the bullish data before it became available.  But the expectation of the lower number was priced in, and the reaction of the actual number became priced in when the data was released.  Since there's no way to know exactly how many points in the dow, or ticks in the MBS market such a report will lead to, the technical analyst instead looks at today's behavior in the context of historical behavior to adjust probabilities of future assumptions proving true.  That's the foundation of technical analysis ("TA")

Even within the realm of TA, there are numerous different studies and analytics one may apply to almost any metric of the market.  Different people might use them differently and with vastly different success.  But one concept that is at home both with the technical and fundamental analyst alike is that of "Range" or "Trend."  These are not the voodoo-witchcraft type of studies where you print today's chart, boil it with eye of newt, drink it, and have visions of the future.  Rather, it's stuff so basic that it borders on the intuitive. 

An example could be the "range-trade" that sometimes occurs in MBS.  For instance, if we look at a longer term chart of MBS in mid 2009, we can see that most of the range lies within 98-26 to 101-00. 

If we combine this technical knowledge with the fundamental knowledge that there is indecision in the marketplace about future direction in the long term, we might surmise that this range has some sort of inertia.  In other words, trading is more likely to persist in this range than to move outside it unless acted on by a force of sufficient significance to move it. With that as a backdrop, we could then soom in to a shorter term view of just this week:

Indeed all major highs and lows this week occurred at or near one of the extremes of the range.  From here we can begin to add additional non-technical considerations such as the illiquidity, choppiness, and historical predisposition of summer time to be range bound.  Then we'd have a longer term range, and a shorter term range, as well as other data to suggest that the range is more likely to hold than break in the absence of significant movement.  By "significant movement," I mean that it is generally safer to plan according to a range until the range is broken by more than a certain amount, for more than a certain time.  Every analytical point of view is different in deciding what those amounts will be.  But for me, I've arrived at a variable framework where I'd like to see prices go at least 3 ticks beyond the trend for at least 3 days (original day of the break + 2 more).  This has prevented early conclusion jumping more times than I can count (ok, well maybe just more times than I can remember).

Whatever the case, with these concepts in mind, if I were considering my pipeline allocations this week, the longer this range persisted, the more likely I'd be to lock when it reached the highs and float when it reached the lows.  Although one can always find reasons to be either optimistic or pessimistic, basic trends should always be considered.  And so it was on the 20th that regardless of the recent positive indications in the treasury market, it was more likely for MBS to maintain the range, or at the very least encounter resistance if trying to mov higher.  When all was said and done, it certainly did encounter resistance, and although it was a better than expected housing print that catalyzed the selling, remember Charles Dow would tell you that price discounts everything.  Looking to the past we see the consistent emergence of trends such as these, and don't think for a moment those days long ago were not subject to unexpected data!

The bottom line is this: When reasonably mainstream technical analysis is applied in conjunction with market fundamentals and trade-flow considerations, we go from stabbing in the dark at complete unknowns to probabilities that can actually leave us and our clients a little wealthier over time.  There will always be surprises and of course today's home's report could have bucked the trend and those who locked on Thursday or early Friday could be frustrated by that.  But that's where Hedge Ratios and your Gut-Flop come into play.  If we crossed that upside and held it for more than 3 ticks for more than 3 days, we'd simply be adjusting our hedge ratio for the conclusion of this trend to the downside as opposed to the downside. 

The side of the exit informs the next range of prices in which new analysis can be applied.  For instance, if this is a firm break to the downside, we'd likely be going back (at least in part) to the longer term information provided by the price levels around 99-00.  Once there, a whole slew of more minute trend considerations can once again be applied, tracked, and contributing to our hedge ratios when they hold or break.  Applying solid logic to the analysis and to hedge-ratios (Gut-Flop) is critical to long term profitability as an originator.  Of course it's predicated on other aspects, but this is one is among the most important.