Weiss report (Why big down day?)

Investment Analysis Clubs Saul’s Investing Discussions
1

Hey Fools,

I read the Weiss report that came out yesterday. It was cited as the cause for yesterday’s massacre and was the focus of the Barron’s article from foodles post #60658. I wrote a detailed summary for my notes and below I summarize the most important points. I am posting this because the report generated a lot of interest on the board yesterday, but nobody seemed to have read it. It felt slightly off topic (macro concerns etc.) so I cleared it with Saul first. I include a link to the report at the end, which includes both the Weiss report and the CIO survey it relies on. Before reading, I suggest reading brittlerock’s post #60676 for a dose of skepticism about these surveys.

Morgan Stanley’s 3Q CIO survey indicates a sharp slowdown in software spend through the first half 2020. Their two-sentence summary for the entire report is: “After four straight quarters of downwardly revised growth expectations for 2019 IT budgets, CIOs expect another 100bps downtick in 2020, with the steepest deceleration in software and hardware. Analytics, security, and cloud computing rank as the most defensible projects.”

They suggest software investors focus on five factors:

  1. Software spend going forward – areas of strength and weakness
  2. Current EV/sales multiples vs historical averages
  3. Current magnitude of pullback vs historical corrections
  4. Second half comps
  5. Lessons from Q2 earnings

My takeaways in each category:

  1. They asked CIOs to categorize IT spend categories into “most” and “least” likely to get cut. The difference between the two indicates defensiveness. The top three most defensive spend categories (least likely minus most likely) are: 1) Business Intelligence / Data Warehousing (BI/DW) (12% - 1% = 11%), 2) Security Software (12% - 2% = 10%), and 3) Cloud Computing (15% - 9% = 6%). The rest are small percentages, not meaningful enough for me to take seriously, with the exception of AI/ML for which the CIO responses cancel out (10% - 10% = 0%) and perhaps Digital Transformation (8% - 6% = 2%). See the link at the end to the report for the rest.

My takeaway is that we are focusing on the right companies e.g. MDB, ESTC, PLAN, DDOG, COUP in the BI/DW camp, ZS, OKTA, CRWD, ESTC in the Security camp, and AYX and PLAN in the AI/ML camp (note overlap). There are others I am leaving out. I think AMZN and MSFT are the main ones in Cloud Computing, which we are not focused on. The report doesn’t include Ad Tech so TTD and ROKU are not mentioned. I am not sure what category TWLO falls in. Other categories with non-meaningful CIO percentages are things like HR Software (PAYC), Collaboration Software (WORK, ZM), Digital Transformation (SHOP? HUBS? PVTL? TWLO?), which to me just means the survey is useless for examining these categories, again see the report for the full list. Note that many of our companies fall into two or more categories. One final note from the report: “deteriorating CIO results give us pause when it comes to legacy software vendors VMW, ORCL, SYMC and ERP software vendors ORCL and WDAY”. (preaching to the choir here)

  1. The gist is that EV/Sales are still ~25% above the five-year average. The historical sector average EV/S is 6.7x. It reached 12x in March 2014. Bottomed around 3.8x in March 2016. Hit 8.8x in September 2018, bottomed at 6.6x in December 2018, then rocketed up to its second peak at 12x a few months ago in June, and today sits at 9.2x. They show a linear regression between 2018-2020 Sales CAGR and EV/2020 sales, it shows the same thing tchalla’s figures show. The most ‘overvalued’ are, in order, VEEV, COUP, OKTA, TEAM, APPN, PLAN, MDB, and ZM. The ones that really stand out are VEEV, COUP, OKTA, and TEAM. ZM does not stand out nearly as much because the sales CAGR is very high. The most undervalued are CLDR, YEXT, PS, and NEWR. Others that are undervalued are TWLO, SMAR, ZEN, CRM, PANW, PD, SPLK. see my notes at the end about an alternative interpretation .

  2. The gist is that compared to March-May 2014, Dec 2015-Feb 2016, and Sep-Dec 2018, we have plenty of room to fall from here, especially for the category leaders. In 2014 the category leaders (FEYE, WDAY, SPLK, DATA, VEEV) fell by 49%. In 2015/6 the leaders (WDAY, NOW, PANW, PFPT, HUBS) fell by 38%, in 2018 (ZS, TEAM, SMAR, MDB, COUP) fell by 8%, today the current leaders (COUP, ZM, OKTA, VEEV, MDB) are down by 16%. (notice a pattern anyone? The pullbacks are smaller each time because the businesses are getting stronger and the market knows it.)

  3. Basically, software crushed it last year so this year is tough … we already know enough about this … MDB, PLAN, TWLO, COUP, and a few others have particularly hard comps

  4. Q2 earnings – we know the story here, market has been ruthless, any sign of slowdown and the multiple gets crushed

My final editorial: One bright spot is that on an EV/NTM FCF multiple, the sector is just about at the low end of its historical range. I think this is what we should focus on. It shows the businesses are improving, generating more cash, keeping EV/FCF low while EV/Sales rises. The aspects of the ‘overvalued’ argument are based on reversion to historical EV/Sales. I think EV/Sales are trending up and will not revert to the historical mean 6.6x. It may get there, like it did in December, but won’t stay there long. Also they include tons of obviously low grade software companies we won’t even consider, which brings down the average. Finally, CIOs expect the share of application workloads that reside in the public cloud to increase from 21% today to 47% by end of 2022, and the share of on-prem to decrease from 63% today to 31% by end of 2022.

The value I see in this report is that it puts some numbers around what we already know and what we’ve already discussed quite a lot on this board - the market is pricing in a slowdown in spend. It’s why we own the best of the best. The companies that we already decided on our own are the best are the same ones this report says should have the most resiliency to a slowdown. The market decided to sell them off, but we don’t follow the market around here, so let’s stay focused on analyzing the best of the best.

The reports: https://we.tl/t-FNIjHvGWli

121 Likes
2

As a growth investor our risk is not market risk per se (unless a real and true bubble) but business risk.

As business risk, who cares about backwards price to sales or such. I don’t care about that figure anymore than I care about last year’s Super Bowl this year. What counts it getting to this year’s Super Bowl.

There is one thing certain in life (as long as you survive) is that one year from now today will be one year in hindsight and two years ahead will only be one year in foresight.

Take Zscaler (again, I am not saying Zscaler over Alteryx or Datadog or any such thing, just using it as an example as I know the exact numbers off the top of my head), it presently has an enterprise value from Yahoo! of $5.45 billion.

Its guidance is $405 million for the current year. If its beats that by 11% it will be $450 million (basically 50% growth, $405 million is 32.7% growth. Thus it does not take much to take disappointing guidance and turn it into spectacular results. Of course growth could also be less (like happened with Nutanix and Nvidia and Talent and New Relic, etc. and thus disastrous). Last year guidance was exactly (plus or minus a point or two) this year’s guidance and Zscaler beat by 17-18%. Will it beat by 11% this year? Who knows, could just as easy miss by 11%. I am investing because I do not think Zscaler will miss by 11% and then Zscaler will likely beat by 11% when all is said and done.

If Zscaler beats by 11%, and revenue ends up at $450 million for the year (vs. $405 million guidance now) the enterprise value to revenue multiple will be 12x. At this time next year Zs’s trailing multiple will be 12x. What will its forward multiple be? 40% growth is not out of the question at all (not conservative but hardly pie in the sky). Forward revenue would then be 1 year from now $430 million.

That puts 1 year forward enterprise value to revenue of 8.65x. AT WHAT POINT DOES THE MARKET THINK DANG, 8.65 TRAILING MULTIPLE WITH THE NEXT YEAR FORWARD, 35% growth (again, hardly pie in the sky) of $850 million or a multiple of 6.41x.

Point being, if you are so concerned about multiples as described in that Weiss report, you have really missed ALL of the relevant facts for a growth investor.

It happens (I cited some above - and Nvidia is recovering, Nutanix, who knows, might as well) so there is risk, always risk. But using the above numbers that I don’t think anyone thinks are pie in the sky or overly aggressive. 1 year from now Zscaler is sitting probably below is cash printing ability if it was a mature company. 2 years from now it is well below that. 1 year from now its 8.65x multiple is 1/2 the buyout multiple of MULE. Materially less than the buyout premium for Tableau.

With all the panic, churn, world is ending, where is the bottom, etc. What we are missing is the risk/reward. All that has to happen is the company you are investing in performs - period. Everything else will take care of itself.

Sure, if Zscaler ends up a great idea that ends up New Relicing, you move on. Move on well before we get to the one year or two year period. But if, as simply an example, Zscaler business keeps on churning on…everything will take care of itself in due course.

That is the value of and promise of category dominant, long-term hyper-growth, disruptive companies. It is about the business performance and everything will take care of itself (absent the usual blah blah black swans, etc.)

Tinker

67 Likes
3

. Good find truewinner.
While this might partly explain the decline I am dubious over the reliability of these polls. And there is this quote "Analytics, security, and cloud computing rank as the most defensible projects.” Most of my Saul type stocks fit in these categories.
Wall Street analysts only think a quarter or at most two ahead, that is how their performance is judged. One who was disastrously wrong about 3 or 4 months ahead but dramatically right about 3 or 4 years ahead would not keep his job long enough to bask in the glory of his long range predictions . Is the same true to a lesser extent with CIO?

1 Like