The tendency to frame the current war in terms of the last one is always great. Each of the last two rate back-ups have led to increased stock market volatility (see rate chart in Figure 1): the back-up of around 40 bps in January
Figure 1: 10-Year Treasury Yields
Figure 2: SPX Index
was followed almost immediately by the late January correction followed by higher volatility through April; another 40+ bps backup in Sept/early October appears to be starting a similar phase as we write. Each time this happens, we observe a rise in "doomsday" articles talking about the rising volume of consumer debt and the coming deflationary apocalypse (for example: Worse Than Great Depression). Inevitably, these authors use nominal debt levels, fail to divide by GDP, and do not take debt service costs and whether those debts are fixed-rate or adjustable into account to arrive at their dire conclusions.
It's Not Consumer Debt
The problem is that people feel uneasy with current stock market levels, especially in certain sectors and names, and associated non-equity market phenomenon that show how "crazy" some speculators have gotten (see Figure 3 for one example). Clearly most equity valuation metrics that value investors
Figure 3: Bitcoin Since May 2016
use have been flashing either orange or red for over a year, so their feeling of unease is justified. Unfortunately, their next step is to look at the last crisis to come up with an explanation for this feeling of unease in terms of some consumer lending sector. The problem is, there is no evidence (Figure 4).
Figure 4: Consumer Debt Levels
|Student Loans||$0.6T||$1.5T||+900B||Rising Income/Long Repayment|
|Auto Loans||$0.8T||$1.1T||+300B||Shorter fixed rate loans|
|Cards||$0.98T||$1T||+20B||Adjustable, high rate|
Source: Federal Reserve, Author Calculations
The mortgage sector, which caused the last crisis, has shrunk by $600 billion and rates are much lower. This means almost everyone has refinanced and housing payments are much lower as a percent of income. In fact, more than enough to offset the small increases in auto and cards payments (remember, there are 150 million+ households in the US, so even $300 billion is just $2000 per household, hardly the stuff of doomsday deflationary spirals).
Student loans are the only debt sector to have ballooned on the consumer side; how much trouble this will cause in the long run depends on how millenials' incomes grow over time as they replace the Baby Boomers in the workforce at middle and upper levels. A last point is that the drop in mortgage debt and the rise in student debt are probably related: high student debt is keeping millenials out of home buying longer, preventing the needed increase in mortgage debt needed to turn millenials into homeowners rather than renters.
On the other hand, as Figure 5 shows, the stock market is definitely a bit out of whack using Shiller PE.
Figure 5: Shiller P/E Chart
Source: Shiller PE
Current levels are higher than than all historical periods except the internet bubble. Other valuation metrics (Altman-Z, market cap to GDP) not based on projections of next year's earnings all say basically the same thing - the US equity markets have been more highly valued only 2-3% of the time.
What's Really Going On?
The story of bitcoin is instructive. Even bitcoin has an attractive story to it - an alternative to gold that has the advantages of anonymity, has nice new-tech properties, sexy math behind it, and uses lots of computing power. Every bubble has, at its heart, some fundamental story that gets it started. The problem begins when the value at which something trades depend only on what someone else is willing to give you and has no external reference to something more fundamental from which the value of that thing is derived. When something begins detached from its fundamentals (like housing did in 2007), more and more people who used to be investors start acting like flippers. Hot money enters the arena, attracted by the rising asset. If there is leverage backing that asset (as with housing) the damage caused when the bubble eventually explodes (as they all do) is much greater.
During the period of QE, there has been a marked rise in the popularity of momentum investing (it's not a coincidence - when gravity is low, pigs can fly). For example, there are no less than 48 momentum ETFs on ETFDB.COM. AUM is now over $20B. There is perhaps another $80B in hedge fund money that follows some type of momentum investing. This is only the "out" momentum community. A number of active mutual funds have become closet momentum investors because they have increased their holdings in the most popular names due to past returns. Examples of top holdings include Amazon (AMZN), Netflix (NFLX), Apple (AAPL), Google/Alphabet (GOOG), Microsoft (MSFT). Passive funds that invest based on market-cap weighted indices also contribute to overall momentum, albeit at a slower-time scale, because each year when the index rebalances, the stocks that have already gone up the most have higher market caps, leading to even more index flows into those names, creating a nasty feedback loop detached from fundamental valuations.
Not all the names I have listed look absurd on a valuation level. For example, GOOG and AAPL look cheap to reasonable to many value managers. On the other hand, head-scratchers include NFLX and the story stock of the century (with a story even better than bitcoin), Tesla (TSLA), which combines a cultish-leader, a world-saving green mission, and a sexy product.
At heart, a stock should be worth the discounted value of its cash flows (Figure 6):
Figure 6: Beware DCF, The Heart of Things
Source: Corp Fin Institute
The problem is that those cash flows are uncertain (at the level of single stocks especially so) and people disagree on what discount rates to use. This leaves the DCF open to being a weapon an overly optimistic person can use to shoot himself in the foot or that a ruthless Wall Street sell-side analyst can use to manipulate the retail masses again and again.
Regardless of what values one chooses to plug in, the important thing is that one can turn the formula around to ask "what earnings growth rates and discount rates are consistent with the current value" to find out what the market "is thinking" (in my career experience, the market spends about 30% of its time thinking and the other 70% wavering between extreme mania and extreme depression, truly an imbalanced dude missing his meds).
What the rate and S&P stock charts at the beginning of the article show is that the market had already priced in a strong economy and hefty earnings growth (and the endless bounty of abnormally high profit margins) but had forgotten about the discount factor, which comes from interest rates (plus equity risk premium). Using the projected earnings of QQQ for example, a 100 bp increase in rates, all else equal, should translate into QQQ being worth about 19% less (this is an easy spreadsheet calculation but you'll have to make some assumptions for the out years) given the very low level of rates we started in.
The other interesting thing that steep rate increases like what we saw the last few days will tend to do is shake out what is most detached from justifiable valuation levels (Figure 7).
Figure 7: S&P, QQQ, NFLX & TSLA
Between Sept 20 and Friday's close before the long weekend (Oct 5), NFLX's beta to the S&P was 2.49 while TSLA's was over 8 (granted, Tesla has its CEO to thank for its excess beta during this time period). Netflix's long-term beta is 1.33 and Tesla's is a measly 1.07 (1-year numbers) despite what you may think from all the noise surrounding the company.
Any further rapid interest rate increases have the potential to snowball the effects shown above. We have a market crowded in a relatively small set of "story stocks" (despite the overall Shiller PE being in the 97th percentile, a surprising 50% of stocks are fair to cheap, which tells you that whatevers got it, GOTIT BAAAD). There are a lot more Mo Mo players in the market than perhaps ever before, many of them not having experienced a true bear in their investing career, feeling secure with their trailing stop loss orders (my helmet has never failed me before so let's see if this bike can hit 200 mph!), all based on the same types of daily-moving-average (DMA) technical analysis or historical beta to the S&P combined with a list of worst day drops.
And they have leverage (Figure 8 just shows retail investors! Hedge fund prime broker lines are not included), the nitro-fuel of people who have not tested their helmets.
Figure 8: We'll Lend You All The Rope You Need to Hang Yourself(aka Margin Debt)
Source: Yardeni Research
What could possibly go wrong?
I'm not saying this could happen next Tuesday (Black Tuesday, anyone?) or even this October. But consider this: The largest single day drop in the S&P in recent history was Oct 15, 2008 when it fell about 8%. Now instead of using 1-year betas, use the betas from Figure 7. How would feel if NFLX fell 20% in a single day and all these stop loss orders all kicked in (the equivalent drop for TSLA would be 64%, but we don't think Musk can do a whole lot more damage, other than by admitting he made up all the financials and skipping out to Argentina)?
Can you picture concentrated flash crashes taking out every single story stock on the same day as computers all whirl wildly in algorithmic epilepsy, foaming at the CPU? Can you picture a smaller broker dealer in the business of selling put options going out of business the very next day because they did not adequately hedge their short put option exposure and underestimated correlations/betas (in a nutshell, CDOs from the last crisis all blew up due to underestimated correlations)?
It ain't about consumer debt, folks. This time it will come from the death of Mo Mo.
Disclosure: I am/we are short TSLA, NFLX. 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.
Additional disclosure: We may initiate additional long or short positions in stock or options on tickers mentioned in the article within the next 72 hours.