Over the past few years running Estimize I've had countless people ask me to write a post on how to make earnings estimates. Before I get into that, let me very briefly explain why this is even important to most traders and investors with a time frame greater than a few days, or anyone holding or trading into and out of earnings.
1. The price of any asset is a function of supply and demand, and those two things are based on sentiment, not pure math. Why do some stocks in the same industry trade for 30 times earnings while others trade for 10 times? Sentiment. A large piece of that sentiment is often the future outlook for the fundamentals of the company (earnings, revenue, margins, the growth of various lines of the business). Estimates on these data points are extremely important, they are one half of the puzzle to price, the other half being the multiple the market gives them (more on that later).
2. Other traders and investors are looking at estimates, which is why you need to be looking at them as well, and understanding what your expectations are given your position in that stock. Just as in technical analysis where any given indicator or trend line is only as important as the number of people looking at it, earnings are a large factor in how most traders and investors make decisions. So if the market is paying attention to it, you better understand the sentiment around what people expect and how they may react to various outcomes.
3. The big boys (Fidelity, BlackRock, Vanguard…) aren't investing for tomorrow, or next week, or even next month. These guys need to pump so much money into stocks to move the needle on their returns that they are looking out 2-5 years. When that's your timeframe for putting money into a company, it's all about expected growth. The timing of their entry into these small/mid cap stocks is often dependent on a large acceleration in that growth, some inflection point that gives them a lot of confidence that they can pump all that! money in now, move the market, pay a premium, and be around to see the meat of the move over the next few years. See Tesla (TSLA) as an example currently of this behavior. Earnings estimates play in to this by being the bedrock of the forecast, because at some point the company does have to live up to a reasonable multiple of earnings, revenue, or some other fundamental metric that the market will attempt to compare to other companies in its industry. The estimates drive supply and demand for the stock.
Ok, enough with that schpiel. I'm going to run through how I make earnings and revenue estimates. Just for quick background, my training was not in building big earnings models in Excel, I've never talked directly to management of any public company about their outlook, and I've never done a proper channel check. This stuff, as I'm about to show you, is completely irrelevant to you unless you are the senior analyst at Fidelity or some $10B hedge fund. My training is in writing algorithms to find a specific set of companies with specific characteristics at a specific time, put together a decent guess regarding where their fundamentals are headed, and then trade them based on the technicals and sentiment on an intermediate term time frame. This stuff is not rocket science, just about anyone can have an informed opinion, here's how.
We're going to use Michael Kors (KORS) as our primary example here. It's a company that's pretty well known, a stock that's being accumulated by large institutions, it's had great momentum, it's growing quickly, and there is a large discrepancy between the sell side consensus and the actual results from the company, as well as the company's own guidance and its performance. All of these things factor into making estimates, so this will be a rich example.
The following things do not have to be considered in this order, but this is generally how I look at it.
1. What is the sell side expecting? Sell side analysts to a pretty decent job at forecast! ing earni! ngs, they just don't do a great job in that last mile, that's where you can add value. The sell side consensus is a baseline scenario that has a lot of inherent bias. But there's a lot of signal there, and they basically get you 80% of the way to your estimate before you look at anything else.
So take a look at where the high, low, and consensus are for the current quarter. If you don't have Bloomberg, go to Yahoo Finance and look it up, here's the page.
(click to enlarge)
(click to enlarge)
So it looks like while the range of estimates for EPS is not that wide, there is large disagreement amongst analysts regarding KORS revenue growth. Most of the time your estimates will fall within these ranges, not always, but most of the time. While the distribution of sell side estimates is often very small, everyone being close to eachother, I would say there is almost always one guy who makes an aggressive estimate in either direction, which gives you a pretty good sanity check regarding what an acceptable range is.
Let's take a look at the Estimize charts now.
(click to enlarge)
(click to enlarge)
As you can see, there is a history here of the sell side being far too conservative with both the EPS and Revenue for KORS. It looks like their margin assumptions have been pretty accurate, they've just far undercut the true growth of the revenue, which has led to their EPS estimates ! being far! too low as well.
We often see this pattern in young high growth companies, and it normally doesn't stop until the growth slows significantly. Given the history here, we would expect that if our growth assumptions aren't terrible, they should crush the sell side estimates again. And unless we believe that something dramatic has changed in the margins for KORS, the sell side calculation should be pretty accurate once again. Personally I have no clue what that number actually is, but it's not hard to eyeball.
We also want to look at the trend in revisions for EPS and Revenue.
(click to enlarge)
(click to enlarge)
A few things are important to remember here. On average, the sell side consensus is about 10% too optimistic 3 months before the company reports. On average, the sell side consensus is too conservative 2 weeks to 1 day before the report. Sell side analysts love to lower their estimates right before the report so that the company can beat them. Now, we don't see this pattern as much in KORS because it's such a high growth company, the sell side is just trying to keep up. In this case, positive revisions are a great sign that KORS is going to put up great numbers again, because if the sell side got even a whiff of bad news, they would be taking these down faster than a freshman at a college frat initiation drinks beer.
It looks like while margin assumptions have not changed through the quarter, the sell side has started that process of tempering their growth expectations as we approach the report date in early August. Given this trend, I wouldn't expect these numbers to move significantly between now and the report unless the company changes guidance or one of its competitors says something interesting.
So that's t! he sell s! ide. It's gotten us 90% of the way to our estimate. But now let's add the last mile.
2. Revenue growth rates are extremely important to stocks showing great momentum. Remember to take into consideration that it's really hard for companies to continually grow revenue at over 50-60% YOY once they hit a certain size. This is just the reality of large numbers. While the market may crush a company that goes from 30% growth to 15% growth in one quarter, it won't necessarily do so for one steadily going from 100% to 90% to 80% to 70% at a steady rate. Let's take a look at the numbers from KORS.
(click to enlarge)
KORS has seen its revenue growth rate decline, but it's still growing at a huge clip given its size. We want to look at a few things here. When it reported a 57% YOY growth rate last quarter, what comp was it up against? Looks like it grew 71% the year before, which was huge. Remember to always look at the comp from the year before, it will give you a good sanity check for the ability of the company to hit those kind of numbers this year. Growing 70% off a flat quarter the year before is a lot easier than growing 70% off a 70% increase the year before.
Then you want to look at what the comp is this quarter compared to last quarter. In this case they are up against a 73% comp vs a 71% comp last quarter. Given that we shouldn't expect a company like KORS to be increasing their revenue growth at this point given their size, we can expect that growth is going to be slower this quarter than it was last. How much slower? Well, the comp isn't that much bigger, so that shouldn't have a big impact on our projections. The sell side consensus is expecting about 37.5% YOY revenue growth. Given that they did 57% against a similar comp, I would say that this is way too low. Do you think that the growth rate really plummeted that much in one quarter? No.
So now we play around with the n! umbers. W! here do we think that growth rate really is. Couple this with the average variance between the sell side consensus and the actual results from the company and we're starting to build a picture. An estimate of $610M for the quarter would put KORS at a 47% YOY growth rate. This seems more realistic.
One more thing to consider regarding growth rates. When it comes to anything retail, you need to take into account how fast a company can build more stores. For a company like WholeFoods (WFM), that's a big issue. You can't just build 500 stores a quarter when you want to, these things take time. There's also the issue of saturation. Understanding where they are in that cycle is important. Also, regarding margins, new stores have lower margins, it takes time for them to get up and running at full speed, so if a company like WholeFoods builds a ton of stores one quarter, expect a dip in margins. The next quarter, expect those margins to recover.
3. Pay attention to the company's guidance. The best and most prolific example is the sandbag job that Apple (AAPL) used to do every quarter. Companies in high revenue growth mode will almost always sandbag guidance because there's really no reason for them to be honest. They are already growing at a good clip, so telling the market they will be at 30% YOY instead of 45% YOY isn't gonna hurt the stock, that is unless it's LinkedIn (LNKD) last quarter, in which case the market had a stupid reaction (see stock at new highs now). The market can be dumb sometimes, it creates opportunities for the rest of us.
We don't show the company's guidance on Estimize, though we are trying to get our hands on this data feed (call us if you have one). Turns out there's really only one firm with a feed I've found. You can look up the guidance in the press release from the previous earnings.
In the case of KORS, they've sandbagged by between $100M and $40M the last few quarters. Will this last forever as their growth slows, no. Will they sandbag big again this ! quarter? ! You bet they will. How much is up to speculation, but again, it's just understanding what the trends in this are that give us more confidence that they are going to beat the sell side estimates. Remember that the sell side often just goes with the company's guidance. This is a large piece of their bias, use it to your advantage.
4. Other margin assumptions are important and you can add value here. Did food prices rise for that restaurant? What did the price of oil do throughout the quarter for that energy company? How much did Netflix (NFLX) spend on content this quarter? These are all things that you can make decent assumptions for to add value. Obviously this factors into your EPS estimate. Do you need to know exactly what the net income is and the number of shares, no, you can eyeball it based on everything else. Remember the sell side will almost always be too conservative here.
Regarding KORS, did they have to discount merchandise or were they able to hold prices steady given demand and the value of their brand? I would say the margins should be fine, I don't see any of this changing this quarter.
5. Use your general knowledge of the company and its industry. I've been super accurate at forecasting WholeFoods because I eat there, A LOT. I've seen their business transform from grocery store to the Starbucks (SBUX) of prepared food. There are huge margin shifts inherent in that transition. Is everyone around you buying an iPad? Well when that thing came out I saw everyone on the subway buy one, immediately. That was a pretty good indicator. These are incremental things that will move your estimates up or down.
How well is the industry doing as a whole? 3D printing is a surging industry right now, so when we are forecasting out 2-4 quarters into the future you would expect that revenue growth is gonna be pretty strong, there's no reason to believe it's going to collapse like the number of devices that BlackBerry (BBRY) is selling. A rising tide lifts all boats.
6. Ho! w did oth! er companies in that industry report recently? These things are often pretty connected. Sometimes it's a zero sum game, as with Coach (COH) and Kors, but a lot of the time it's not. This is going to be a good quarter for all of the home builders, they didn't eat each other's lunch, they all killed it. Use the previous reports from other companies as a proxy for where expectations are and where the companies came in, it's usually a pretty good guide.
Estimize has sector and industry pages so that you can check all of the stocks in one industry by report date. Here's consumer apparel:
(click to enlarge)
7. Take a look at where the Estimize consensus numbers are relative to the sell side and the actual results historically. The Estimize consensus is more accurate than the sell side about 70% of the time historically. There's no shame at all in using this information, look at the delta in the numbers for the current quarter, use that info to your advantage.
8. Take a look at where the very accurate Estimize analysts are with their estimates. This is why we have the scatterplot chart, so that you can see who the outliers are, and if one or more of them has historically been very accurate for that stock or industry, you're going to want to take that into heavy consideration when making your own estimate. This was the KORS scatter from last quarter.
(click to enlarge)
9. The last thing I look at is how the stock has performed over the course of the quarter, and how it sets up technically into the report. As fundamental expectations are a major driver of supply and demand for a stock, you can expect that if analysts are taking their estimates down all quarter, there's a good chance the stock is gonna get smacked. Likewise, if owners of th! e stock a! re selling en mass, you can expect that there's probably something wrong going on fundamentally. This isn't always the case, but more often than not it is. Use the price action as a judge for where to expect the numbers, the buy side buys stocks faster than the sell side moves their estimates.
This doesn't always give you an accurate picture of where the company will report, but it will give you insight into what the expectations are and how much of those expectations are baked into price already. Google (GOOG) just had a hugely disappointing quarter, estimates were real high, and they blew it.
Put all of this information together and you've got your estimate. Again, this stuff isn't rocket science. The media likes to make this whole thing sound really complicated, it's not. I've been a buy side analyst, we traded tens of millions of dollars worth of stock on a daily basis, we never had any earnings models in Excel. Let others do the 80% leg work for you, then add the last 20% intelligently, because that's where the real money is made. It's not made in doing the calculation from net income to EPS, it's made in having a good view that the company you're looking at is going to grow revenue faster over the next 4 quarters than the rest of the market thinks, and being able to understand how the market may revalue its multiple because of that. This is the true value of an analyst, it's not in crunching numbers, any fool can do that given a month of training at an investment bank.
Storing your estimates inside of the Estimize platform can be a valuable process given the alerts and notifications that it produces. This dramatically cuts down on the time it takes to do the 9 things I just laid out above, and you can incorporate that information in real time as things move.
If you have any other questions please feel free to post them below, or get in touch with me directly.
Disclosure: Nothing on this site should ever be considered to be advice, research or an invitation to buy! or sell ! any securities, please see the Disclaimer page for a full disclaimer.
No comments:
Post a Comment