PGMI index

The PGMI indicator combines corporate earnings, growth, reasonable valuation, predictability, and inflation to provide a useful overall assessment of market valuation.

It works by quantifying, on a weekly basis, the number of S&P 500 companies that meet a set of criteria. These criteria act as a filter to identify companies worth considering as investments based on their predictability, earnings growth, and reasonable price. In this way, the number of qualifying companies tells us how favorable or unfavorable market conditions are for investing at any given time. A high number of qualifying companies signals that the market is cheap and that we can overweight equities; a low number signals that the market is expensive and that we should reduce exposure or exit altogether — either because corporate earnings have deteriorated significantly or because valuations are stretched and unsustainable. Markets also have their own momentum, which is why a specific time window for acting is defined.

The indicator allows us to identify bear market periods driven by earnings depression, price bubbles, and undervaluation periods — and tells us how to respond in each case.

You can get the list of companies that pass the filter here: PGMI filters

Technical Description

First. Normalized EPS figures (source: TIKR.com) are used for each S&P 500 company, as reported by the companies themselves. Using the EPS from the last 4 fiscal years, growth (k) and predictability (r2) are calculated by fitting the following expression (h and k are fitted):

e_n = h · (1+k)^(n-1)

Second. The number of years to recover the investment (Y) is calculated based on normalized earnings, factoring in growth and inflation. Y is derived from the following expression, where I is the inflation value (using the T10YIE rate from the Federal Reserve Bank of St. Louis) and PER is the current Price/EPS ratio:

1 + k · PER · (1+I)^Y - (1+k)^Y = 0

Third. Filter used to calculate N, the following criteria are applied across all S&P 500 companies to count how many satisfy all of them simultaneously:

• PGMI Filter
  • k > 0.1. Growth greater than 10%. Below this threshold, the company is not worth analyzing on a growth basis.
  • r2 > 0.95. High predictability (r2). Below this threshold, earnings can no longer be considered predictable.
  • Y <= 12. Investment payback period under 12 years (CAGR > 7%), which represents a reasonable minimum return threshold.
  • Sustainable over 3 years. Analyst-estimated EPS for 3 years out must exceed the most recent fiscal year figure. If not, recent growth is considered circumstantial and not expected to continue.

You can find more detail here: Investment Plan for Companies with Predictable Earnings

How It Works

• Weekly, after Friday's market close
  • If the number of qualifying companies N is greater than or equal to 50, we can overweight our market exposure, and will do so for the next 180 days. This signals that, based on corporate earnings and their growth, the market is cheap. Caution: we must assess the market from other angles to check whether there is an overriding reason for the market to appear cheap — such as a credit or commodity crash triggering deep and sustained declines. We should avoid overweighting in such scenarios. The 200-session and 200-week moving averages attract significant institutional activity and can serve as a useful reference to observe price behavior around those levels.
  • If we are not within a 180-day overweight window, and N is greater than or equal to 40, we remain invested — this will be the typical case. We stay invested for the next 180 days.
  • If N is below 40, and we are not within any 180-day window, we simply stay out of the market. At this point the outlook is very poor — earnings are depressed or the market is severely overvalued.

The chart shows the grey windows indicating periods when we should be out of the market, and the thick green line marking overweight periods.

Risk Considerations

Overweighting must be approached with great caution — by default, it is better not to do so.

• Overweighting Risks
  • It can happen that when the overweight signal appears, the market is in a sharp short-term downtrend. In such cases, extra caution is warranted.
  • You should analyze the correlation within the index: if it is very low, the broader market may be in an uptrend while the index falls, because market conditions are unfavorable for the companies that dominate it.
  • There may be cases where an earnings bubble in the market creates the false impression that stocks are very cheap, when in reality those earnings are unsustainable.

Future Development

I am currently exploring the option of using EPS without NRI figures from GuruFocus with a rolling years calculation, so that the methodology becomes more standardized and signals can be generated well ahead of those derived from fiscal year data.

You can get the list of companies that pass the filter here: PGMI filters