Are we in another Bull Market now?

Update 8/18/2020

Well, that didn’t take very long! On August 18, the S&P 500 and Price Index finally set a new record high. The Total Return Index (dividends reinvested) had already reached its new all-time-high on August 10. So, we know for sure that this is a new Bull Market and it started after the March 23 Bear Market Low!

2020 Case Study - update 8-18
S&P 500 returns, total return and price index since the 2/19/2020 peak.

May 11, 2020

The stock market is still well below its February all-time-high, but it’s holding up remarkably well considering how poorly the economy is doing right now:

  • Double-digit unemployment,
  • 20+ million jobs lost according to the BLS payroll employment data,
  • 30+ million jobs lost according to the sum of weekly unemployment claims since March,
  • -4.8% GDP growth in the first quarter (annualized rate) and possibly a 10% drop quarter-on-quarter in Q3, which would be reported as a 34% drop annualized!
  • And a whole host of other economic indicators that look so bad, you’d have to go back to the Great Depression to find similar readings

Actually, “holding up” is a bit of an understatement when you look at the stock market as of the Friday, May 8 close:

  • The S&P 500 TR index rallied 31% since the March 23 low!
  • We’ve recovered more than half of the drop peak to trough!
  • Since April 8, we’re no longer 20% below the recent all-time-high, which is often quoted as the cutoff for the Bear Market!
2020 Case Study
The 2020 Bear Market up to May 8. Did a new Bull Market start on 3/24?

Does that mean the Bear market is over now? I certainly hope so! Whether the Bear Market is over and a new Bull Market might have started already obviously depends on the definition of Bull vs. Bear. What is that definition of Bull vs. Bear anyway? Different people have different definitions, some more sensible than others. And it gets more complicated: even if we do agree on a sensible definition, there could still be uncertainty over what’s the state of the market right now because some criteria cannot designate the state of the world in real-time but only after the fact. There’s a confirmation lag! So, we could be in a sort of a bull/bear-limbo state right now! How is that possible???

So many questions! Let’s take a look at how I think about Bull vs. Bear…

Bear Market Definition

Let’s start with the slightly easier of the two, the bear market definition. That’s often defined as follows:

“A Bear Market is a drop of 20% or more from a recent Bull Market High.”

For example, the 2020 Bear Market was confirmed on March 12 when the S&P 500 dropped by more than 20% relative to its recent all-time-high. (Side note: the Dow Jones already reached that cutoff a day earlier.) In any case, do you notice a subtlety here? There are two distinct dates to consider:

  • The Bear Market started on 2/20/2020, the first time we fell below the 2/19/2020 all-time-high.
  • The Bear Market was confirmed on 3/12/2020, the first time we fell below the -20% cutoff.

Thus, between 2/20/2020 and 3/11/2020 we were already in a bear market, but we didn’t know it at that time. In other words, …

We were in a Bear Market Limbo between 2/20 and 3/11!

The reason we were in that limbo state was that if the market had recovered from the drop before hitting the -20% and had then reached another all-time-high above the 2/19 level, the Bull Market would have been intact and nobody would have ever talked a Bear market again. Similar to the experience in the fourth quarter of 2018. But, alas, the market did drop below the magical -20% threshold and we confirmed the Bear Market on March 12.

Bull vs Bear Limbo State
Haven’t fallen more than 20% from the recent Bull Market High? We’re in a Bull/Bear Limbo state (top panel). If we reach a new all-time-high later without breaking the -20%, the whole episode is still a bull market (middle panel). If we do break the -20% mark, then the entire drop peak to trough will be a Bear Market (bottom panel).

It’s hard to wrap your head around the fact that not everything can be confirmed in real-time. Especially today where all the pertinent information is supposed to be at your fingertips and we’re used to overnight-shipping and even same-day delivery.

But this kind of data uncertainty is all around us. I’m an economist by training and the analog of this type of data headache is the issue of timing the business cycle turning points, i.e., the beginning and end of recessions and expansions. In the U.S. that’s done by the NBER (National Bureau of Research). The NBER doesn’t declare a recession at the first sight of some minor economic weakness. We have to see enough of an economic contraction before a recession is confirmed. And the starting point of the recession will be that earlier date when the weakness began, not the date when it was finally confirmed.

Just like pregnancy is confirmed when the pregnancy test comes back positive. But the pregnancy started at the time when, uhm… well, you know when, right? 🙂

Sometimes this limbo-state really boggles your mind. For example, the 2001 recession started in March 2001 but the NBER didn’t make that announcement until November 2001, which, incidentally was already the end of the recession. And that end of the recession wasn’t declared until July 2003. I can see how people have a tough time wrapping their heads around it. But if you’ve ever done any sort of serious data science you’ll know plenty of examples like this!

Side note: Actually, the NBER faces (at least) two additional challenges over and above the Bull/Bear timing: First, there is no fixed criterion like a 20% drop in stocks. The NBER does not use the often cited naive recession definition of two consecutive quarters of contracting GDP. Rather, the NBER looks at multiple monthly economic indicators. It’s mostly up to the discretion of a committee of Econ Professors to declare what they deem a sufficient contraction. Second, macroeconomic data are delayed and subject to revisions. Especially the Payroll Employment numbers are often revised months and even years after the initial release. Contrast that to the Bull/bear timing. The all-time-high on 2/19 was visible to everybody at the time of the market close and it will never be revised. The (hopefully) Bear Market low on March 23 will never be revised either, so people in finance have much easier data to deal with. So, economists have to be a lot smarter than the finance folks to deal with all those data pathologies! 🙂 

Let’s now move on to the even slightly more complicated Bull Market definition…

Bull Market Definition

So, what would confirm a new Bull Market? Two definitions I’ve seen floating around recently:

  1. The market “leaves the bear market territory,” i.e., we cross the Bull Market high minus 20% again.
  2. A 20% rally from the most recent Bear Market Low

The S&P 500 actually satisfied both criteria recently: On April 8, 2020, the S&P 500 Total Return Index came within 20% of the old all-time high (“leaving the Bear Market Territory”). And on the same day, you also saw a cumulative 20% return over the March 23 low. So, both definitions have recently been used in the popular media calling the end of the 2020 Bear Market. 

WSJ calling the end of the Bear Market
The Wall Street Journal called the end of the bear market on March 27 when the DJ Industrial saw a 20% rally vis-a-vis the March 23 bottom. This is a screenshot someone sent me on Twitter. Incidentally, I could no longer find this article on (but please send the link if you can find it). Maybe the WSJ realized that their Bull Market was a bit loose???

Unfortunately, the two definitions don’t make a whole lot of sense. It’s because both criteria would have given you nonsensical Bull Market starts during previous events. Case in point, the 2000-2002 bear market, see the chart below

  • The 2000-2002 bear market had a 22% rally between September 2001 and March 2002. But that shouldn’t have qualified as a new Bull Market. The 2000-2002 episode is normally considered one single Bear Market event rather than two bear markets with a short Bull Market in between.
  • The “leave the Bear Market Territory” definition is even more problematic. In the Spring of 2001, we “left the bear market territory” because the market rallied and came to within about 13% of the old all-time-high. Was that a Bull Market, then? If so, there would have been a new Bear Market starting in May 2001 with a new Bear Market low in September 2001. But then the Spring 2002 recovery would have again come within 20% of that May 2001 Bull Market high. So you would have had at least three separate bear markets straddling two very short bull markets. Not a very sensible Bull Market definition, because the 2000-2002 stretch is normally considered one single event!
2001 Case Study
In the 2000-2002 bear market, the +20% and the “leaving bear market territory” criterion would have generated false bull market signals!

So, let’s try another criterion:

The Bull Market is confirmed if we reach a new all-time high. (and again: the Bull Market would have started at the trough, not the day it was confirmed)

Well, that’s certainly a sufficient but not a very good necessary criterion. For example, following the 2000-2002 Bear Market, we didn’t reach another all-time high until 2007, right before the Global Financial Crisis. Does that mean we were in a limbo state for a total of 5 years between 2002 and 2007 while the market rallied quite substantially? Would anyone argue that if the Global Financial Crisis had arrived a little bit earlier without reaching a new high, that essentially the entire decade was one long Bear Market? Probably not. The “all-time high” criterion is probably a bit too restrictive.

Another example of the “all-time high” criterion being too limiting: The market didn’t reach its 1929 all-time high until the 1940s. But it’s commonly assumed that 1932-1937 constituted a new Bull Market even though the 1937 high was still slightly below the 1929 level.

In a Twitter discussion a while ago, Dr. Leif from Physician on FIRE linked to an article in Fortune with a better Bull Market criterion:

“a 20% advance off of the prior low that does not get undercut within six months

I actually like this definition a lot because it eliminates all of the past temporary bear market retrenchment (to my knowledge) but also allows for a reset and a new Bull Market high even if it’s below the previous all-time high (e.g. 1932-1937).

So, using that sensible definition, we still find ourselves in the dreaded limbo state right now. We cannot say for sure if we’re currently in a new Bull Market or if we’re still in the Bear Market that started in February. It all depends on how the market moves from here, see the chart below:

  1. Bottom Left: If we drop below the March 23 low over the next few months, we are still in the same ugly Bear Market.
  2. Top Right: If we bounce around for the next roughly 5 months and we get to October 8 (= April 8 + six months) without a new low, then we are indeed in a new Bull Market, even if we haven’t reached a new all-time high. And any new bear market after that would be considered a separate event. Either way, we’re going to be in this limbo state for another close to 5 months, until October 8. Ugh!!!
  3. Bottom Right: If we reach a new all-time high in the S&P 500 we would consider the entire path from March 24 to that day a new Bull Market.
Bull vs Bear Limbo State2
In a limbo state gain (top left panel). If the market tanks again with a fresh Bear Market low, then forget about that Bull market! This would be considered one long Bear Market (bottom left panel). If we don’t see another low for a total of 6 months, then we are currently in a Bull Market. Any fall below that Bull market high would be considered a new Bear Market. Or, we get to a fresh all-time high, in which case, the path from 3/24 to that high is all a new Bull Market (bottom right panel).

Why is the stock market holding up so well in the first place?

Obviously, we can yap and yap about the Bull vs. Bear but that’s all semantics. The more pressing question, of course, is why are people talking about a Bull Market in the first place if the economy is so awful? I’ve written about it in previous posts, see the “Reasons for Optimism” and the “Financial Lessons from the First Quarter” posts. I also still stand by the points I made on the April 20 ChooseFI podcast “Have we seen the Bottom?” But I may write another update on my thoughts in the future. Stay tuned! 🙂


There’s a good chance that we’ve entered a new Bull Market starting on March 24 this year. But we’ll have to wait a little bit longer for the final confirmation. That doesn’t sit really well with a lot of people who have to know the information right away. People are so impatient these days, especially journalists! But it’s the nature of the beast, that the sensible bull vs. bear definitions all have a confirmation lag built-in. Bull/bear markets, recessions/expansions and pregnancies all have that in common: a confirmation lag.

There’s nothing we can do about it, so just sit back and enjoy the ride. Decades from now, I will tell my grandkids about how I navigated the great 2020 Bear Market and the subsequent Bull Market. Even though I had no clue in real-time when the Bear ended and the Bull Market started.

Hope you enjoyed today’s post! Looking forward to your comments below!

33 thoughts on “Are we in another Bull Market now?

  1. It’s actually not the entire stock market that is holding up so well. It’s driven by a few stocks such as Microsoft and Amazon – you could say the S&P`s top 5-10 – that are heavily weighted and have contributed strongly.

    1. Excellent point. But isn’t that true every time? There’s always a sector (not always the same but always 1-2 sectors) that do better and take the index out of the bear market while the rest of the index is hurting.
      Another reason to do indexing! 🙂

  2. Ern, let me go out on a limb and say the scary unemployment number would be below 10% by November.
    Let us see if my bullish prediction made on 11th May 2020 holds 🙂

    1. Haha, that’s a very optimistic forecast. Keep in mind that the current reading, 14.7%, is for April and that survey was conducted around April 12. It’s possible that
      that in May we still go above 20%. That number will be released on June 5.
      I’d be happy if we’re under 12% unemployment by November. Below 10% would be even better, of course.
      But I agree: By November everything will look much better already! 🙂

  3. Very interesting and apt thoughts Karston. I’m just wondering why there isn’t a rule that says “20% drop off of the prior high that does not recover within six months” for a bear market to qualify as one.

    It’s like pouring a soft drink from the can into a glass too fast, the foam gets too far ahead and even almost spills over the glass. You wait a few seconds for the foam to disappear and then keep on pouring. At no time were you losing anything of substance. The “bear” was an illusion.

    1. Yeah, that’s a good point. There should be an additional confirmation criterion on the down-side. I’ve seen people apply a criterion like that and claim that 1982-2000 was one long Bull Market. So, they filtered out the 1987 event, exactly for the reason you brought up: It didn’t last long enough.
      So, long story short: Even for the Bear definition there are different definitions.

  4. To add to the confusion some might look for the price of the S&P500. Some might look at the inflaton adjusted value. Some might look at intraday values and some at the end of day data.

  5. Great piece.
    I’d be interested for your thoughts as to when we are safely out of the ‘threat’ of a double-dip. In reference to both the tech-crash and the credit crunch, there was a ‘suckers rally’ (or whatever ridiculous name you fancy) that really hurt people before it got better. Without googling, I believe that was rather prolonged and quite out the range of impatient journalists and some bullish chief economists.

    1. Good point. That risk is still there. My fear is that if politicians keep dragging their feet and resist opening we’ll certainly see another downdraft.
      But the stock market seems to put a low (not zero) probability on that.

  6. Interesting analogy about confirmation lag for pregnancy ! As a doctor, I would like to add other nuances, for the fun of it. Pregnancy doesn’t actually start with a couple’s “quality time”. It starts at conception, which can happen up to 5 days later. Then when you get to calculating the pregnancy timeline, you actually calculate the number of weeks based on the date of the start of the last period. So conception occurs typically 2 weeks later, but that can vary wildly, because after their period, some women take more time to ovulate than others. It gets very confusing, which is why we appreciate having dating ultrasounds later on in the pregnancy to get a better idea of the due date.

    1. Ha, I had a suspicion that a smart MD will find a way to correct me there. I’m glad I kept the language vague. The “you know when, right?” could also refer to that, haha!
      Thanks for letting me know! 🙂

  7. I do wish we’d stop focusing on the U3 number and pay more attention to the U6. The broader view given by the U6 is, in my opinion, a better indicator of the job market and the health of the economy. Given that the U6 is creeping near 30% (including estimated losses since mid April) and that layoffs are continuing, I foresee economic pain and low earnings for at least the next three quarters. Consider also that the impact of the shutdown on public sector spending and employment will take much longer to play out than the private sector, and I think it’s Pollyannish to expect recovery by the end of this year.

    1. Unemployment (and Underemployment = U6) are lagging indicators. So, you’re right, there’s more pain in the labor market. Expect 20+% in U3 in May. But the peak in the UER normally occurs several months after the Bull Market started.

  8. That is if your/our grandkids will bother hearing about the stock market at all. The way things are going with Universal Basic Income and other socialist tools capitalsim will slowly die.

      1. If you stay away from SPY and only trade ES and SPX you come under the CFTC regime which is federal not state. Also you can charge 2/20% for any investor. 😉

  9. So if we are looking at symmetry in definitions, a Bear market would be ” a 20% decline from the prior hire that does not recover within six months”. Since the high was Feb 18 and the new high Aug 18, by that definition the Bear market never happened and we are still in the longest running Bull market of all time.

    1. Hah, that’s a good point. For that reason, I would not be surprised if some analysts will call for retroactively renaming this a bad correction.
      But then again: there was a recession, so I fully support the idea of calling it a bear market! 🙂

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.