To: Haim R. Branisteanu who wrote (343 ) 8/10/2003 9:42:32 AM From: Jim Willie CB Read Replies (1) | Respond to of 110194 Economic Cycle Research Institute (ECRI) misses refi's ECRI's Weekly Leading Index (WLI) does not include the refi mortgage component of the Mortgage Application Index (MAI), which is regularly published by the Mortgage Bankers Association. The WLI only includes the purchase mortgage component of the MAI, so it underestimates the strongly negative impact and implications of the important leading indicator importance of refis and their recent tremendously ominous decline and the similarly negative effect on overall mortgage applications. For instance, the analysis and commentary from ECRI fails to note that the MAI is a whopping 48% off its spring highs with the refi mortgage component dropping 32% last week alone, compared to the only 3.5% decline in the purchase mortgage component. And not only does our work indicate that the MAI is bound to decline further, but as pointed out in my late-April e-mails on this subject, we don't expect the spring peaks in the MAI to be exceeded for decades, which is consistent with our previous Great American House-Equity Bust call. By now everyone knows how important refis are to consumer cash flow and spending, which kept that most important sector (roughly two-thirds of the US economy) from joining the recent recession that was almost exclusively led down by the business sector. The recent collapse in refis is another mark of the end of consumer spending strength (see our comments on recently flagging chain store sales), as well as our expectations that it'sa significant fundamental cause to such an incipient slowdownand second recession. Since I'm using the words "slowdown" and "recession", I will deviate slightly from the main points I'm making and point out why we believe serious investors should be suspicious of the use of the politically-correct concept and word "recession," especially as utilized by perma-bulls, new bulls and the thus influenced financial media. A Kitchin business cycle has four distinct stages: recovery, expansion, slowdown and contraction. (See Burns and Mitchell, founder of the NBER, 1946 book, Measuring Business Cycles. However, it is more politically convenient and thus popular in the financial media to refer to the business cycle as having only two phases: expansion and recession. In the politically correct version of referring to the business cycle, a recession includes only one of the four business cycle stages, "contraction,"while "expansion" asymmetrically includes the other three stages, including the "slowdown" stage. Of course, the stock market disagrees as it almost always declines during the "slowdown" stage, giving rise to the economist's maxim about the stock market's economic false positives, that the stock market forecasts four out of every three recessions. Now back to the leading economic indicators. ECRI points out that six-month growth rate of their WLI, now at 11.5%, has risen for 14 consecutive weeks and is at its highest point since the late 1980's. This is a reference to what amounts to a leading indicator of a leading economic indicator. This is analogous to looking at acceleration as a leading indicator to velocity, which itself is a leading indicator of the economy. Looking at "leads to leads" can be very dangerous if one is not very analytically careful and, as I've reported before, ECRI is much more bullishly biased in this and other pick-and-chose statistical references of they are wont to present, than they are analytically careful. (I do not make this charge lightly.) First, they fail to note the very high volatility of what is effectively a second derivative of the economy, or more accurately the first derivative of some leading indicators of the economy. The high volatility of their arbitrarily selected six-months growth rate of their WLI makes it much more difficult to use as a forecasting tool for the economy. Instead of their six-month rate of change high-volatile oscillator, the first attached chart illustrates how we use a four times larger, or 104-week (two-year), moving average to better smooth out the oscillations in our modified version of ECRI's WLI and tomatch the NBER's belated recession calls in the business cycle, but our index works in real time. Our modified WLI essentially adds the NASDAQ to correct for the NYSE-only bias in the WLI. In the second attached chart we demonstrate the inaccuracy of ECRI's claim that the recent recession was similar the 1990 one. We think the increased severity this time resulting from both a deeper and longer correction in the WLI is an important precursor to the unsustainability of the current economic rebound and the resulting coming second recession. All of this is consistent with our expectation that the widely touted economic rebound following the first recession is both predictably failing and that it will soon be followed by the second of what will likely be four (give or take one) rolling recessions during this K-cycle winter, or what we quantify as a deflationary economic BAAC Supercycle bear market period - explained in our SMECT report:safehaven.com Bob Bronson Bronson Capital Markets Research