**Introduction**: Founded in 1931, *The Value Line Investment Survey* (VLIS, also Value Line or VL in context) has seen it all in the stock market. I began subscribing to it in 1979 and continue to be a subscriber. VL is not easily fazed, and I've only seen it be either bullish or relatively bullish on common stocks. It has developed a predictive formula based on three variables that estimate the average price of the Dow Jones Industrial Average ("DJIA" or "the Dow") for the year ahead. It regularly published this prediction at yearend. The Dow is now perhaps 2000 points above what this formula suggested would be its average price for the year, even though corporate profits and dividends, two of the three variables, are roughly in line with consensus to date.

Because this comes from a perma-bull organization such as *Value Line*, investors may wish to read on. Investors who rarely change their stock allocation may find this particular article of little interest, though.

The culprit per this VL analysis: The interest rate upturn this year.

The VLIS formula, if adjusted for the current level of interest rates, is now predicting an average level for the Dow this year well below current levels. Of course, the Dow is where it is almost halfway through the year for its own reasons, and the lack of higher interest rates to compete better with stocks may continue to make stocks appealing. And of course, the economy could speed up considerably, which could help shares.

As most of us know, the best-known tracking ETF for the DJIA is the SPDR Dow Jones Industrial Average ETF (NYSEARCA:DIA).

**Background**: As stated, at yearend VLIS publishes a forecast for the average level of the DJIA for the coming year. They do this via a mathematical formula, which they tweak each year based on the prior year's information. It has worked very well for years, though in 2008 and (very likely) 1999 it failed to predict the excesses in each direction. What is good is that it is empirical, ferreting out what has actually been predictive of stock prices, as opposed to imposing its views on what "should" happen based on "fair value" or other concepts.

Their philosophy in this regard is exemplary, I think.

Last year the DJIA's average price was about 1/2 of one percent within that which VLIS predicted at the end of 2011. Each of the past four years, VLIS has been reasonably accurate with its yearend forecast. Please note that they are not predicting the Dow's level a year in advance; they are predicting the yearly average. Here's what they do, in general terms.

**The VLIS formula (general concepts)**: There is one constant and there are three variables to this formula. Again, this is an empirically-derived formula based on what has "worked". First, VLIS adds a 2% constant to the prior average level of the Dow. Thus last year's approximate 12,700 average Dow would predict, if the three variables remained unchanged, about 13,000 for this year as an average level for the Dow.

The three variables they have determined are the relevant ones are year-on-year dividend growth, interest rate changes for AAA-rated longer-term corporate bonds, and earnings growth. Interest rates are a positive for this formula when they are *falling* and a negative when they are *rising*. Naturally, rising dividends and profits are both positive for stock prices in their system.

Nowadays, dividends on the Dow are quite predictable, so the major unknown variables before the year starts are interest rates and earnings.

The yoy changes are subject to exponential transformation rather than being simple ratios.

**VLIS 2013 estimates**: At the end of December last year, VLIS projected an average Dow for 2013 of a little over 13,300. This would have suggested a rise to this level by midyear and a close over 13,600 by yearend, if stock prices happened to march monotonically upward throughout the year. Such as result would have been fine for the Dow, though not stellar.

This was based on the following assumptions: moderate dividend and profit growth, and unchanged corporate bond interest rates. Here's the problem, from Bondsonline.com, showing the 10-year AAA corporate bond rate throughout this year up to June 8:

If we assume that the VLIS estimate for growth of DJIA dividends and profits is accurate for the rest of the year, and that interest rates do not change at all the rest of this year, I'm going to make a guess that its formula would project that this about 25% increase in rates (from about 2.20% to about 2.75% -- about a 0.55% absolute increase) would approximately equal or even outweigh the combined bullish effects of the anticipated (consensus, more or less) dividend and profits growth. This would leave its projected average Dow at around the 13,000 that multiplying last year's average price by the 1.02 constant would suggest.

**Implications**: Without doing any math, because this is all back-of-the-envelope stuff, let's say that the average Dow this year so far has been around 14,250. Let's say that this represents 43% of the year, with about 57% left. If the Dow were actually to average 13,000 or so this year, that would, quite roughly, suggest that it should average about 12,000 from now on.

A 12,000 average price for the rest of the year implies a range above and below that. So that would be a disaster for the market. This number assumes normal economic growth.

For the Dow to merely come within 7% of 13,000 would be a big deal.

**The "other" Value Line analyses**: A different analysis that VL provides weekly is a projected 3-5 year "appreciation potential." As of last Monday, this was at a near-record low of 40%. While that sounds pretty good, VL cautioned a few years ago that typically it overestimates returns by a full 5% per year. So, a 3-5 year projection of 40% gains is more likely 20%. On a 4-year time frame, that's 5% per year. Value Line's 3-5 year price projections for individual stocks include dividends, so that's what its analysts implicitly are suggesting as an annual total return for the Dow. This datum for the entire market has been very good in my long history with VLIS. On July 13, 2007, VLIS gave stocks only a 35% projected total return 3-5 years out. On March 9, 2009, it estimated a 185% 3-5 year total return. Similar negligible appreciation potential was signaled near the market top in 1987, and so on.

**Summary**: Value Line has developed a rating system involving three variables that have predicted DJIA action in the year ahead with good accuracy, though with significant variability from time to time. Of these variables, two have the greatest level of unpredictability: interest rates and earnings. Using the presumption of unchanged AAA corporate bond interest rates this year, this formula predicted an average Dow of about 13,300 for 2013. Given the sharp rise in rates so far, stock prices may be ahead of themselves even with the benign, consensus economic climate that Value Line expected.

Also, and separately, company-by-company analysis of stock price appreciation potential created by VL's analysts suggests that the general stock market is near its lowest level of projected 3-5 year price appreciation ever, at least so far as my knowledge goes (back to the mid-1970s).

**Cautions**: While it's obvious, I'm putting forth this info and analysis because Value Line is independent and has a steadily bullish agenda. When I average the predictive formula discussed first, and the cautious appreciation potential analyses discussed later, and then think of alternatives, no clearly good ones pop up that would merit asset allocation to them. So it may well be that the "no alternative to stocks" meme continues to carry the day, overvaluation concerns even from perma-bulls notwithstanding.

Another caution is that the DJIA is (obviously) different from the other U.S. stock market indices, which may have faster growth ahead and may prospectively be superior investment vehicles.

**Final comments**: Value Line, which makes its living selling stock market information and advice, has now published two different sorts of data that raise elevated downside risks to stock prices even without an unexpected economic slowdown. This strikes yours truly as at least somewhat different than similar analyses coming from a short-seller.

As discussed above, the VLIS predictive formula flags substantial percentage rises in interest rates as dangerous for stock prices. The absolute level of rates is not pertinent to this formula. Of course, rates have not been this low for many years, so perhaps this parameter will prove to have been irrelevant, though I'm not ready to accept that viewpoint yet.

Will past be prologue in the recent 50+% rise from last year's lows around 1.4% to the current rate of 2.16%? (While the absolute level of 2.16% is very low, it is about a point above the yoy Consumer Price Index change.)

As a partial answer to that question, I'm mindful that Ben Bernanke said the following in summing matters up in last month's testimony to Congress:

Recognizing the drawbacks of persistently low rates, the FOMC actively seeks economic conditions consistent with sustainably higher interest rates. Unfortunately, withdrawing policy accommodation at this juncture would be highly unlikely to produce such conditions.

So he may be nervous about whether the economy can easily handle a rapid increase in rates. And if he's nervous, I'm nervous also.

That consideration is good enough for me to take the VLIS analyses seriously in my thinking and have a little more spare cash than usual hanging around should this steamroller of a bull market retrace some of its steps at some point not "too" far away from now (either in time or in price). Longer term, though, I continue to agree with Value Line and many others that well-priced, high-quality common stocks (or truly cheap lower-quality shares) are indeed fine vehicles for most individuals, and that making drastic asset allocation changes based on one or two models or formulae, or even several of them, often works out in the fullness of time to be an error.

**Additional disclosure:** Not investment advice. I am not an investment adviser.

Comments()