Description

Buy Oracle (ORCL)

Summary

Oracle is an attractive and defensive long.  

The main reason to own it is because Oracle’s growth is finally accelerating after going through multiple years of complex business model transition to Cloud and subscription. Oracle Cloud Infrastructure (OCI) is gaining legitimate traction with marquee client wins (e.g. Uber, Zoom, Department of Defense etc.) and operating margins should inflect upwards with scale efficiencies.  Of course, existing Oracle customers are also migrating to their cloud products 

While other cloud companies and software companies are showing decelerating revenue growth, ORCL’s overall revenues and cloud revenues are accelerating (expecting 30% organic cloud rev growth in fiscal 2023)  and management is guiding for continued acceleration in 2023.  And we all know that Wall Street loves accelerating revenue growth (which Jim Cramer calls ARG), which usually leads to multiple expansion as well.  Especially in this environment, it’s hard to find an accelerating revenue story, particularly in tech.  

Oracle cloud has a niche for sophisticated and large companies who are looking for cheap and fast bare metal / IaaS, and don’t need all the bells and whistles that the other cloud providers have.  Also, it is a sneaky AI play as well, as many AI startups and companies actually use Oracle cloud and prefer its speed and lower cost.   In NVIDIA’s most recent earnings, they specifically called out Oracle twice as of their partners for NVIDIA DGX loud 

I believe Oracle is an attractive software Cloud investment in an economic downturn/recession scenario as well, given its focus on enterprises. Its products are mostly B2B and mission critical and locked in for the vast majority of their customers (e.g. Enterprise Resource Planning, Human Capital management, etc.).  Oracle’s products are more ‘back office’ and hence less macro sensitive than ‘front office’ type products 

It’s core database and broad suite of software products given them a lot of diversification

Also, ORCL’s management have been good stewards of capital, reducing share count,  and have made accretive and generally decent acquisitions.  Consistently improving their margins over time

Current expectations are still fairly muted, and valuation is attractive vs. peers.  

Oracle only grew organic revenues and EBIT at a 1% CAGR from 2012-2022,  so clearly people thought it was a dinosaur and value trap.  But in that time they have been building their cloud business, buying other software companies (accretively) and reduced share count by 45%

Organic revenue growth constant currency is now 9% and they’re projecting that to continue in the medium.

Oracle is also projecting operating margins to expand back to pre-pandemic levels of ~45%, +300bps by Fiscal 2026.  The Cerner acquisition and cloud is weighing on margins at the moment.  By then Oracle will be 50%+ Cloud, and will be viewed more favorably

This leads to 12% EBIT % CAGR and mid-teens EPS % CAGR. 

Even Oracle’s software business is doing well,  and their Autonomous Database is growing too 

Oracle trades at 16x 2023 P/E vs. peers at 20-30x P/E for now similar levels of growth.  Looking forward, I think 2024 EPS will be up 15%+ to $5.80, and I think the multiple can expand to 20x, leading to 30% upside (plus a 1.4% dividend yield).  

Regardless of whether or not they’ll get multiple expansion, it’s a solid compounder from here, and still relatively underowned and disliked.  I don’t think the multiple will contract too much from here, and there’s not much downside to revenue or earnings in a macro downturn given Oracle’s usual mission critical product, back office focus, and usually recurring revs.   Oracle’s database, despite being viewed negatively by investors and the market, is still the leading and fastest relational database and is not going away. 

In a downside case, I think maybe it could go to 14x on $5.60 in EPS +1.4% dividend = $80.  -12% from here. 

Oracle Cloud

Ellison poached a bunch of top cloud execs from competitors to develop Oracle Cloud.  In some ways, coming in later has helped since they could learn from the short-comings at the other cloud companies.  And also, they were tasked to make this positive margin from the beginning despite being at lower scale. 

Oracle Cloud is a legitimate option.  Oracle has been rising up Gartner’s magic quadrant  https://www.gartner.com/doc/reprints?id=1-2BFV0YLG&ct=221019&st=sb

Their advantage arises from a few factors.  Because they are newer and don’t have as many features, that makes OCI less buggy, cheaper, and faster.  While AWS, Azure, and Google Cloud rush to add features and functionality for everything and whatever users request, Oracle is focusing on the basics.  Hence it’s a cheaper option for sophisticated tech companies and large companies.  

Oracle Cloud overview: https://www.youtube.com/watch?v=-xZPijlAIKg

The way they’ve architected the data centers (which they lease to minimize spend and improve speed to market) and internal networks with ethernet has helped the speed and cost efficiency of OCI.  E.g. the way they’ve made it so each server’s memory is rapidly accessible by other servers has helped improve the speed and efficiency (Remote Direct Memory Access - RDMA) - and hence it has been an attractive compute option for customers.  Unlike other clouds, Oracle Cloud also has much lower data ingress & egress fees.  These fees have been very annoying for customers at competitors. https://www.oracle.com/database/technologies/exadata/hardware/rdmanetwork/

Price comparisons depends on a lot of factors, but many organizations have indeed found OCI to be cheaper in certain use cases than competitors:  https://www.oracle.com/cloud/economics/

From what I understand, overall Oracle Iaas/Paas is much lower gross margin (30%) but still EBIT positive.  

It has many well known customers, and was even part of the US Department of Defense cloud contract along with the other 3 big cloud companies.  Many of the large automakers use Oracle, so does FedEx, Zoom, Exelon, Uber, etc.  https://www.oracle.com/cloud/customers/.  Toyota uses them for computational fluid dynamics and to store geospatial data for car navigation.   Fedex uses them to manage their logistics fleet, and uses analytics and database as well. 

https://www.oracle.com/cloud/service-comparison/ 

Combined with Oracle’s expertise in databases, their core compute and storage options are actually quite good. 

They recently won a big deal with Uber Feb,  it’s a 7 year deal with Oracle and Google (totaling over $1bn), with Oracle winning the majority of it (for the core server compute requirements.  Uber is also using Oracle ERP and database for parts of its business (freight) 

Anyway, the proof is in Oracle’s numbers, total and cloud revenues are accelerating, and margins are holding in well with cost management offsetting cloud margin pressure.  So they’re doing something right and customers are eating the cooking

Only 11% of Oracle’s customer base uses Oracle Cloud, so there is a lot of low hanging fruit to pick up still 

There was also recently an Information article detailing how AI companies find Oracle’s faster and cheaper. 

Oracle Winning With AI Start-Ups: 

https://www.theinformation.com/articles/ai-startups-find-an-unlikely-friend-oracle

The Information reported that Oracle is an improbable early leader in the race to be the cloud provider for the burgeoning group of AI start-ups, even though it has <5% of the cloud market. The article says at least six venture-backed AI startups, including Character.ai, Lamda Labs, VizSeek, MosaicML, and Adept AI Labs, primarily rely on Oracle for cloud computing. It also highlights that two founders of AI startups told The Information that Oracle can run complex machine-learning models more economically than Amazon Web Services or Google Cloud. Oracle benefits from having configured its server hardware—specifically, the cables that connect processors to one another—to drive down prices for firms running large machine-learning models. For example, in late 2021 when MosaicML paid sticker prices while testing all of the major cloud providers, Oracle was consistently 23% to 33% cheaper. In addition, the process was much faster on Oracle: Training the model took roughly 90 minutes on Google Cloud, 70 minutes on AWS and 60 minutes on Oracle. MosaicML ran similar tests involving another image recognition model, ResNet-50, and found Oracle was anywhere from 50-80% cheaper than Google Cloud, and 25-50% cheaper than AWS (interestingly MosaicML said it couldn’t measure Azure costs because Microsoft wouldn’t provide on-demand cost data). 

The piece notes Oracle took a different route than peers and uses ethernet cables to connect GPUs to each other. Oracle says its investment in the ethernet cables allows it to run AI models using fewer GPUs, driving down costs (Ethernet cables are also typically cheaper than Nvidia-designed cables, according to MosaicML’s Frankle). < Larry Ellison has touted this as one of their advantages too.  

 It’s still early days in the AI boom but this article seems to dovetail with Larry Ellison’s comments back in December that AI business was “exploding” for Oracle.

Investment Thesis in more detail

Oracle is entering in a new growth era with accelerating growth. I believe it will be viewed more as a “Cloud” company than a mature software company

-         After 10 years of low revenue and operating profit growth of 1% CAGR, ORCL is finally seeing its cloud business reaching critical mass. As a result, its revenue and EPS should accelerate to HSD-LDD% range going forward

o   At its October 2022 Analyst Day, ORCL laid out ambitious (but not believable) long-term targets calling for $65bn in revenue and 45% non-GAAP operating margins by FY26, representing a 9% revenue CAGR and >300bps of margin expansion from FY23E levels <although some of this is in-organic

o   Oracle Cloud (IaaS and SaaS) is expected to account for ~50% of revenue and surpass database revenue by FY2026

§  Oracle Cloud is expected to grow at ~25% CAGR for the next several years. It will nearly double its Cloud business from F23-26, adding $15bn in net new revenue  

§  I think once ORCL reached this 50% tipping point, it will help broaden its peer comp universe, and make the stock more appealing to SaaS/cloud growth focused investors  

·        Oracle Cloud revenue: FY2021 22% (pre-Cerner), FY2022 26%, FY2023 31%, FY2026 50%+

·        Number 1 or 2 SaaS business in the world by revenue

o   Includes leading strategic SaaS offerings such as Fusion HCM – human capital management, Fusion ERP – enterprise resource planning, and NetSuite which account for ~55% of total SaaS revenue. Expected to grow 20%+

o   Remaining ~45% of SaaS revenue comes from vertical applications including Cerner. Expected to grow LSD%

·        4th largest public cloud business behind AWS, Azure, GCP

o   This diversification narrative is still in the early innings, since only single digit percentage of their customer base adopted Oracle Cloud thus far

§  ORCL has seen good momentum in its cloud business. It just announced a 7-year strategic cloud infrastructure partnership with Uber on 2/13. This award is estimated to be over $1bn TCV. Uber was previously an AWS customer

§  Gartner has ranked Oracle's Cloud Infrastructure ahead of Google's Cloud Platform. As Oracle continues to publicize deals with big logo names (i.e.  Zoom, Uber) and gain recognition, it will help drive adoption of its Cloud services

§  Oracle Cloud is favored by some customers because it is competitively priced and lower frills, which appeals to some customers. 

o   As of FY22, Growth, Stable and Declining revenues are 22%, 54% and 24% of total software revenue per Bernstein estimates.  The mix has improved over the last few years. ORCL’s revenue has stabilized and inflected higher

Recession resilient business model

-         Oracle’s business is more recession resilient than other software companies, for a few reasons

o   Oracle is likely to benefit from consolidation as enterprises look to simplify their tech stacks in a downturn

o   Oracle’s business is all about supporting critical workloads for the largest businesses in the world. These workloads are going to hold up well even under increased inflationary pressure    

§  The shift of ERP to the Cloud should accelerate as companies focus on driving cost saving while delivering better management of their businesses

§  Oracle Autonomous Database automates out the need for very expensive IT staffing which would help customers in inflationary conditions and provides better security and access control

§  Oracle OCI Gen 2 has unique value propositions which should help it in theory sustain growth

Upside to margins as Oracle archives more efficiencies post the Cerner deal

-         ORCL saw significant margin erosion after the Cerner acquisition

o   The Company should be able to get back to 45% operating margin territory by FY26 (from 42% in FY22 levels) as its cloud business gains more scale, drives more cost efficiencies (from R&D, S&M, and G&A), and get more cost synergies at Cerner

o   ORCL is going to move Cerner into Oracle OCI, and replace software that Cerner currently purchases/rents from other software companies with Oracle software

-         Combined continued share buyback (they shrunk share count by 45%after a decade of aggressive buybacks), which drives more EPS growth and supports the stock

Valuation & Comps

ORCL is trading at an attractive valuation given accelerating growth outlook.

 Historical valuation are less relevant given ORCL’s new growth prospects of HSD% revenue, DD% operating income, and mid-teens% EPS growth

ORacle is trading at 16x 2023 P/E,  while now putting up similar growth to its large cap peers (peers trade at 21-30x 2023 P/E); I think that multiple disparity will shrink. 

It’s trading barely above its long term historical average P/E as well,  so people are still not believers, which is fair since Oracle’s history admittedly hasn’t been too exciting and management likes to overpromise and underdeliver.  But this time, their progress is real. 

Historical Valuation Chart

10 year NTM P/E

 20 year NTM P/E

Channel Checks

To summarize, recent partner channel checks, they’re actually overall net positive across Oracle’s segments, eg.g. cloud, ERP, HCM, etc. with most partners meeting or exceeding targets and sounding pretty positive

Catalyst

- ongoing cloud growth and acceleration

- more marquee customer wins

- execution of mgmt targets (which their compensation is highly tied to)