Blog Archives

Importance of Risk Management

9/13/2023

The goal of the Satori Fund is to be an “all weather fund” that can protect capital or generate positive returns even during severe market declines. As such, I believe in good RISK ADJUSTED returns. This is one of the reasons why our Fund was able to produce positive returns in 2022. Additionally, and for the record, we are up this year (2023) including the months of August and September (month-to-date through 9/12) while the S&P 500 finished negative in August and is down again in September.

One of the metrics that we use is the RSI (Relative Strength Index.) To oversimplify, if this gets above 70 or “overbought” as defined by the index, we will think about trimming a position.
If a stock that is overbought also reaches a price level that has acted as a temporary ceiling in the past, this will increase our likelihood of trimming.
If the stock has had some fundamental development recently, resulting in a surge in the stock price to overbought levels (e.g. new product launch, positive update to business trends, positive earnings release, recent bullish presentation by company management) this will also increase our likelihood of trimming the position. In effect, the news or event we were waiting for has now occurred and is priced into the stock. Apple (AAPL) product launches are a good example. AAPL is up nearly 70% of the time in the month prior to a product launch with an average gain of over 3% but then is typically down the day of the event.
If multiple wall street analysts that cover the stock have also recently upgraded a stock or raised their price targets after the stock has elevated to an overbought level, this is also a warning sign for us to trim the name.
If valuation is high relative to the history of a stock, this is another warning sign that any pullbacks or disappointments could be more severe than normal.
If there is an earnings report coming up, and the stock is overbought heading into the event, we will trim an even larger amount of the position than normal or even all of it. We did this even in front of the recent NVDA results even though it is probably the hottest stock in the market right now. We will typically reduce an oversized position into an earnings event unless we have extremely high conviction AND it is an out of consensus view.
If the earnings event happens and it is fundamentally not what we expected, as long as there is enough after-market liquidity to be able to sell the stock at a level that is not considered oversold, we will do so. This is not only good risk management but also helps us clear our head and make better decisions. We are not married to a position just because we own it already. A great example is our sale of META in the after market on 2/2/22 despite the stock being down ~15%. It closed the next day down 27%.
If the position is oversized (over 10% of the portfolio in our case) and it is now overbought, this is also a good reason to reduce the position size even if there is not fundamental news event coming up.

INTC - Intel

The stock reached an RSI of 74 during the day on 9/11 following the news that they were selling a stake in IMS to TSMC. The stock intra-day was up 51% YTD.
This follows the CEO giving a positive update to earnings guidance earlier in the month and announcing a prepaid customer for their foundry business.
The position grew to be one of the largest in years.

Our plan with all trims made to a position due to risk management reasons and NOT due to fundamental reasons, is to buy some of it back when it is no longer oversold and back at an interesting technical level. This is the case again with INTC which we trimmed yesterday.

ORCL - Oracle

The stock reached an RSI of 71 the day of their earnings release and was up 55% YTD
Two major Wall Street firms upgraded the stock in the two weeks prior to the earnings release
The position had grown to ~10% of the portfolio

We sold most of our ORCL position on 9/11 prior to them reporting earnings and hated doing it given our belief that it would be a beat and raise quarter (we were wrong.) But we try to stay disciplined to our risk management principles above which have served us well over the nearly 20 years of running the fund.

When ORCL, reported revenues a touch below consensus, we sold the rest of our position in the aftermarket down roughly 5%. This slight revenue miss was fundamentally NOT what we expected. Then on the conference call later that day they guided both revenue and EPS below expectations for the upcoming quarter and the stock went down further. Finally, yesterday (9/12) the stock closed down 13.5%.

Our risk management principles, though we hated doing it at the time, saved us from losing much of the substantial gains we had achieved. In investing, it is a batting average and over-time, our risk management principles have improved that average.

At this point, our plan is to do more research including data from ORCL CloudWorld next week before deciding what our next move will be on the stock. At a market PE of 20x, ORCL is appealing to us versus the 32x PE for Microsoft (MSFT) for what we still believe should be somewhat similar earnings growth over the next few years. We still believe that ORCL is becoming a credible #4 cloud provider behind Amazon ($AMZN), Microsoft ($MSFT) and Google ($GOOGL) but Oracle’s recent guidance below the street expectations for both revenue and EPS for their upcoming quarter does not fit well with that belief. In the technology industry especially, things are always changing and for us, facts always trump beliefs.

“When the facts change, I change my mind - what do you do, sir?”
― John Maynard Keynes

2 Comments

Market Thoughts for Barron’s Interview

9/5/2023

1 Comment

Summary

We became short-term bullish on August 18th
Our favorite stocks are AI related including $NVDA $INTC $GOOGL $AMZN $META $ORCL.
Student loan repayments restarting in October and “excess savings” being depleted will make Q4 more challenging
Inflation, the Fed & Bond Yields may stay higher for longer than expected
The S&P valuations is on the higher side at 21x versus 19x when inflation has been at these levels
Our fund is meant to be an “all weather” fund. The fund was profitable last year with the S&P total return down 18% but also this year with the S&P up 19%. We were also up in the month of August despite the 2% decline in the S&P.

Recent Events

On August 18th, we posted how we were covering over half our shorts and expected a short-term bounce based on the bond market stabilizing. In fact the S&P did bottom on 8/18 at 4370 and 10 year yields peaked the next day at 4.34%. This past week saw a 4 day rally in the S&P through (8/30) after having no consecutive up days in the month of August.
https://twitter.com/DanielTNiles/status/1692581872625398214?s=20

The S&P is up 3% from its close on 8/18 with Nasdaq up 6% while 10 year yields have now dropped by 17 bps from 4.34% at their peak to 4.17%.
Jerome Powell’s speech on Friday 8/25 was marginally positive followed by data of a slowing labor market this past week while PCE readings were inline with expectations

On Tuesday, JOLTS showed job openings at their lowest levels since March of 2021 on Tuesday along with lower than expected consumer confidence
On Wednesday, ADP said private employers added 177K jobs this month vs 200K expected
Core PCE on Thursday was inline with expectations at 4.2% though up from 4.1% in June
Bond yields after reaching levels last seen in 2008 on August 19th, dropped sharply this past week on lowered expectations of the Fed needing to raise rates

Stock Picks

Our favorite large cap technology names with an AI angle to them.

Nvidia (THE AI enabler in chips) & 15% below its historical PE for its forward 12 months
Intel (foundry biz gaining an AI customer like Amazon or Nvidia before year-end and trading at just ~1.4x book value excluding MobileEye)
Google (#3 AI cloud enabler with more information than anyone else for training AI and trades at a 24x PE)
Amazon (#1 AI cloud enabler & retail profitability ramping)
Meta (AI beneficiary for video recommendations & ad placements and trades at just a 22x PE)
ORCL (becoming a credible #4 cloud infrastructure competitor and at 22x PE, it trades near a market multiple of 21x vs Microsoft at 31x)

$NVDA beat revs by 10% in their April quarter and guided 53% above the street for their July quarter when they reported on May 24th. This was the fuel for a surge of 8% on the S&P from ~4200 to ~4400 from (5/24-8/23) with NVDA up 54% during that time and Nasdaq up 10%.

The AI optimism remains alive and well as NVDA on August 23rd beat revenue consensus for their April quarter by 21% and guided revenues 27% above the street for their October quarter. Since Nvidia’s latest results, the S&P has gained 2% from ~4400 to ~4500 from (8/23-9/1) with NVDA up 3% during that time and Nasdaq up 2%.
The good news is that Nvidia’s stock on a forward twelve months basis is actually ~15% below its average PE over the past 5-7 years as estimates have gone up faster than the stock. We still think we are in the early innings of the ramp in AI but obviously at some point in 2024, supply will catch up to demand and the double and triple ordering we are seeing from customers today will have to normalize.

But just because every technology company says AI 50 times on their earnings calls does not mean that all companies are going to benefit. “In the year of efficiency” to borrow a term from Mark Zuckerberg, spending on AI is in many cases coming out of spending in other areas. We would note that most big ad agencies have missed their revenue forecasts. Most big networking related companies (such as Juniper, Commscope, Nokia, Ericsson) have missed (with the exception of Arista.) Many big IT services vendors have missed like Accenture, IBM and Infosys. And now even software security seems to be feeling the pinch as seen by Fortinet results. Even in semiconductors, TSMC, the foundry for Nvidia, had to cut their numbers while AMD also saw an estimate cut. Even Microsoft which is a big shareholder in OpenAI which developed ChatGPT saw their Azure estimates get cut slightly. Market multiples are too high for stocks not to feel pressure if this continues.

We think our top AI picks will be able to withstand the economic pressure between now and year-end on discretionary spending

Student loan repayments for 40M restarting after being on hold for 3 years.
~$1 trillion in excess savings (vs $2.5T at Covid peak) will be nearly gone by year-end.

Inflation, the Fed, Bond Yields & Valuation

With solid US GDP growth, higher energy prices and a still solid labor market, inflation & therefore the Fed and 10 year treasury yields may remain higher than investors expect:

The easiest inflation compares of 9.1% peak CPI in June of 2022 are behind us.
Oil has surged from a low of $67 in mid-June to over $80 today putting upward pressure on the most important commodity for inflation.
Wage growth remains at 4.4% with 50% more job openings than people unemployed and unemployment at a low 3.5%.
Japan moving their YCC target to 100 bps also puts upward pressure on global bond yields. Their 10 year yields are at the highest levels since 2014 at 0.65% and were flattish to negative from 2016 to 2021 which acted like an anchor on bond yields across the globe.
The Fed also has to issue over $1 trillion of Treasuries in Q3 and $850 billion in 4Q23 as it rebuilds cash reserves from the debt ceiling crisis and supports greater than expected gov’t deficits. This should put upward pressure on bond yields.

19x is the average multiple for the S&P over the past 70 years when headline CPI is between 2-3% vs a 21x multiple today with the CPI at 3%. So valuations are on the higher side.