This article is a follow up to my previous piece on crude oil refineries. You should read that piece first to understand that this is a continuation of that story which showed how refineries looked to me at the time and how they look now. You can read the article here.
I am a full time investor and I have the time to analyze both the macro and micro in the oil business. While I don't work in the industry, I have a passion for understanding how it works and I follow fellow author Robert Balan. His work is right here on Seeking Alpha. Here's a link to his blog.
I am learning on my own how to be a better independent investor. I will admit that this is taking a lot of my time, so I have found it helpful to write out my thoughts, share them with fellow seekers here on SA and this accomplishes two things. First, it clarifies my thinking as writing down a thesis forces me to see if it makes sense. But almost as important are your comments that allow me to see where the thesis is weak or just offer me a different perspective that I hadn't considered. Believe me, there are no dumb questions, all questions provide value and I try to answer them all if I can. Ideas are my currency and I'm happy to read other ideas.
Starting with the macro economic environment, SPY is down 3% since August. Many stocks including oil companies and refineries are down in 2016. I expressed in September that refineries were showing value and that maybe that was a good time to enter.
I recommended Holly Frontier (NYSE:HFC) and Western Refining (NYSE:WNR) at the time. HFC was $23.85 at the time of writing and today it's $24.44 or a gain of 2% vs SPY with a decline of nearly 3%. WNR was $27.69 at that time and today is $30.65, a more respectable 22%. Many people may feel that's the end of the gains.
What Do The Models Say
Balan has introduced an updated refinery model.
Click here for a live link.
This is an updated model which does not use synthetics. Synthetics helped make the models easier to read at a glance, but avid users preferred the older ones that showed the three model lines separately. It can help investors understand what forces affect a commodity or equity rather than simply showing the model result.
It appears that the rise in the earlier model has shifted ahead in the updated model. This is common in models that the trends and math are correct but the timing can change. These are time-variant models so time itself is a variable and the further out in time you go, the less accurate they become. They are also multi-variant models which means that many time-variants go together in the math which means that if the rates of change used in the models change just a little bit on each component, the models themselves can have compounding errors. It's not that the models are wrong as they often work out eventually, but you have to understand that when time is a variable that every second matters.
I have come to trust the models but understand that time is variable and may not appear exactly as modeled. That doesn't mean they don't help you, but you need to understand exactly how they can help you. If you bought a refinery based on my article in September then you have a gain of either 2% or 22%. This is common. Company specific performance can vary even in the same macro environment. HFC was unloved and trending down and this company specific feature may continue.
Note the models show a short term dip which I would consider a buying opportunity should it occur. It's important to see the upward trend for the year.
Events In The Sector
CRAK is the refinery ETF that I also discussed. It was $19.76 on the date of my first article and today it is $20.04 or +1.4%. So in that sense, both of my picks did better than the sector but the industry did not react any differently than these companies. It's important to always compare your stock pick to the sector.
Here's an annual regression of CRAK. An annual regression is how the entire sector has performed this year.
This shows is the performance of the ETF CRAK and the solid blue line shows the trend which is flat on an annual basis but rising since August. An annual regression encompasses all of all the highs and lows for the year, so even though the sector hasn't changed the ETF is actually expensive on it's annual routine because of the rise since August. The center line is the annual regression trend, the light red line above means one standard deviation expensive and the darker red line means two standard deviations above the mean. This means expensive on an annual routine. The green lines represent the same delta to the cheaper side of annual regressions.
It's more useful to show the ETF CRAK against the general market SPY. It is in this regression you can see that the refinery sector is less expensive on it's annual routine vs SPY and is has trended down, but since August is now trending up. This does not mean the sector is cheap as these regressions are relative to each other, so refineries are relatively cheap in an expensive market. I see this in the valuations and the dividends as well.
So in a down trending market refineries are higher relative to the market but the absolute prices are much less expensive than earlier in the year.
Here is the annual regression of WNR vs SPY. You can see that it's current price is much higher compared to the market than even the refinery sector represented by CRAK. So while the trends are up and certainly must come to an end at some point, VAR models are still showing refineries are going up.
So if refineries are relatively expensive but cheaper on an absolute level, what does this mean? You cannot compare relative performance to absolute performance directly. You have to understand that if the market goes up less than refineries than refineries are relatively expensive. What happens if the market goes up and refineries go up less? Then refineries will become relatively less expensive even while stock prices on them rise.
Technicals show technicians that the sector is expensive on an annual routine and some may not even try to ride refineries higher. But investors with technical insights react differently. We look at valuation first before we look at the macro because expensive stocks will earn less over the long run if we pay too much. Value investors always look for value. Since we cannot value a sector, let's take a look at VLO, another large refiner in this sector. I wrote about HFC because it seemed good value to me, but you can see that WNR performed much better. What other refineries might we consider if the model is accurate and the sector is due for a rise? VLO is very large and well known player in this sector. Here's it's technical chart. You can see it is also in an uptrend since July, similar to WNR. VLO is still making higher lows although top line price growth has stalled.
And here's it valuation using FASTGraphs.
While the PE is actually higher than normal, but not compared to the market, the strong dividend shows you the relative undervaluation. When crack spreads increase, profits will increase. What affects crack spreads is gasoline demand and crude costs. Low oil prices actually help refineries and the market has priced in much higher oil prices in the face of a weak economy it only makes sense that profits will fall and they have. The investing thesis on refineries has to include lower (relative) oil prices and higher gasoline demand. In that environment refineries will do well. What's important to understand is that North American refineries have access to the majority of shale oil due to proximity. Refineries commit to oil contracts and buy oil far in the future using hedges. The hedges that E&P buy fix the selling price of oil to the refiners. So while oil prices will rise in the future, US based refiners have a whole many months worth of raw material already priced at $50. When oil rises and gasoline prices rise around the world, US refineries will have the lowest cost and potentially the highest profits we have seen in a while.
A high dividend payer with a positive outlook will appeal to an income investor. That's where I see VLO in this mix. Not as cheap as HFC, but a larger refiner with access to seaborne grades which help make it an flexible and efficient refiner.
WNR By Comparison
WNR has outperformed HFC handsomely. But let's look at this mid continent based refinery like HFC. From a technical perspective it is trending up like VLO. However it is slightly more expensive than its 50 day MVA but the moving averages are showing a golden cross, a technical signal that traders will like. Rising volume confirms the breakout.
FASTGraphs confirms that WNR is slightly more expensive than VLO compared to history, but is still priced lower than future earnings.
Refineries continue to show value to me. Income investors with technical insights have a different view than pure technical traders. Valuation and dividends matter to the price of a stock and VAR Models show how the drivers of refinery profits are improving. Not straight up, but improving and that is a constructive macro environment that readers of this article now know ahead of the crowd. I am choosing to add to WNR and VLO at this time. I will continue to hold HFC at the current time, but VLO and WNR are trending up more.
Disclosure: I am/we are long WNR, VLO, HFC, HEP.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.