PGT Industries (PGTI) is an attractive cycle play on the Florida housing market, is far and away the cheapest building products biz in the market, and has ~50% upside over the next 6 – 12 months. Since PGTI’s recent August 2015 highs, the market has penalized the stock down from $16 to $9 – 10 / share given increased uncertainty around the duration and magnitude of the Florida housing cycle. The market believes demand is tapering whereas I believe PGTI has >3-years of runway ahead to get back to “normalized” historical levels + PGTI is opportunistically growing inorganically through accretive acquisitions that effectively give PGTI a dominant position within all of Florida (over the past couple years through two acquisitions, PGTI has effectively diversified away from being highly reliant on Southern Florida).
MARKET VIEW: FL housing market is stalling given: (1) interest rate concerns, (2) foreign buyer concerns (i.e. currency moves are damaging foreign buyer interest) and (3) perception that the FL housing "recovery" is over (i.e. 2010 - 2015 was the recovery); MY VIEW: FL housing still has 3 - 4 years of "recovery" ahead given single family units are still ~50% below "normalized" historical levels (~65k units versus normalized of ~100k) and housing affordability is still attractive providing a significant demand tailwinds for the business. More notably, housing starts in FL are currently still 2-standard deviations below 1980 – 2003 average levels (i.e. excludes the most recent boom/bust).
Divergence: As context, for the past ~2-years, PGTI stock has treaded water (~$8 – 9 / share in mid-2014 and currently ~$10 / share). In the meantime, (1) PGTI has made two accretive acquisitions that effectively give PGTI dominant market positioning in FL (~$210MM in size / ~40% of the TEV) and (2) there is increased evidence of the housing recovery.
PGTI is a $500MM market cap business with a largely reactive sell-side base. The buy-side appears to be pricing in an “okay” 2016E and then petering out. On 2016E figures, my estimates are ~11% above on EBIT and ~18% on EPS. An interesting way to back into “what’s priced in” is the following exercise: assume NO growth in housing starts in the out years (i.e. ~65k housing starts / ~2 standard deviation level below average historical but let’s assume this persists) – this would suggest PGTI trades at a ~10% FCF yield on 2018E metrics. This analysis suggests the market is not pricing in much additional new construction growth
Evidence supporting divergence: (1) FL single family units are still 2-standard deviations below 1980 – 2003 average historical levels - this equals a >55% increase to get back to normal; (2) Relative to the US at large, FL housing (a) witnessed a painful drawdown (87% fall-off from peak to trough … versus US average at 67%) and (b) still furthest away from "normalized" (~55% upside to normal versus US average at sub-40%) = this provides more runway to revert back to normal; (3) While there are valid concerns (foreign buyers / FX squeeze + interest rate risk), FL single family permit activity is a good leading indicator and continues to flash green (~20% YTD through February 2016); (4) Affordability analysis below highlights rates are a risk, BUT still far from being a pressing risk
Valuation / New Reality: As the market grows more comfortable with the duration and magnitude of the housing recovery (10%+ organic growth), a premium multiple will be placed on PGTI. Currently, PGTI trades for less than 10x PE, >10% FCF yield and ~6.5x EBITDA-capex on 18E estimates (a more "normal" environment).
BASE CASE: Applying a more reasonable multiple of 14x PE to normalized gets to a >$15 / share FV, and implies a ~7 - 8% FCF yield and ~9x EBITDA-capex. This scenario assumes a 4-year recovery to normalized (100k units) so uses a slightly-below normal earnings power estimate given 2018E as the valuation point
UPSIDE CASE: Assume a 3-year recovery to normalized (~120k units) and apply a 14x PE multiple = $20 / share
DOWNSIDE CASE: Assume an overly-punitive scenario in which FL housing remains at 65k starts (i.e. 2-standard deviation levels below normalized) and apply a 10x PE multiple = $8.25 / share. Assuming a slightly less punitive multiple of 12x PE would generate a ~$10 / share FV
I do not hold a position with the issuer such as employment, directorship, or consultancy. I and/or others I advise do not hold a material investment in the issuer's securities.
Better results than expected (or simply less bad than feared), additional accretive capital allocation (share buybacks), improved visibility on FL repair and remodel and housing starts