Editor's note: This story was originally written going into 2023 and has been added to our CNBC Pro white paper vault. In my 15 years of covering the economy, I've learned a thing or two about business cycles, and the recessions that end them. It's amazing how much confusion there still is about these concepts, which are so central to our lives. But with these pointers in mind, hopefully you'll feel better prepared to know when we're actually in a recession, when we're not, and when we're about to be in one. Let's go in order: 1 | THE "YIELD CURVE" The first place where recessions usually show up is in financial markets, in a thing called the "yield curve." There are several variations on this, but it's all the same central theme: Is the yield on short-term Treasurys (like 3-month bills, or 1-year notes) higher than the yield on longer-term ones, such as the benchmark 10-year? If so, the yield curve is "inverted," and savvy investors start to panic. This is a very reliable early sign of recession. The problem is it tends to happen a year or more before the actual downturn begins. So, you have to put up with, sometimes years, of mind-numbing debate about whether to ignore the yield curve (my money is on not) or whether we're actually going into recession (which we typically do). 2 | STOCKS, COMMODITIES AND HIGH-YIELD BONDS The next warning sign is typically a sell-off in things like stocks, commodities and high yield bonds. You have to consider these together so you don't get misled. In fact, to be in recession, you typically have to have all of these pieces lining up. If they don't, don't jump too quickly to any apocalyptic conclusions. But if you have a bear market in stocks —typically defined as down at least 20% from their peak — and a similar drop in commodity prices, that's never a good sign. One way to confirm their message is to check the trading action of high-yield bonds, using something like the iShares iBoxx High Yield Corporate Bond ETF (HYG). If this is also dropping, it means people are nervous junkier companies can't repay their debt. Just one word of caution: this can be a bit more lagging than stocks, commodities and the yield curve. Don't be surprised if it takes longer for severe stress to show up here. 3 | ISM MANUFACTURING INDEX AND NEW DURABLE GOODS ORDERS Now, let's move on from markets to the "real" economy. Early warning signs of a downturn often show up in the ISM manufacturing index. When it slips below 50, it indicates the sector is contracting, and it's a bad sign for the rest of the economy. Watch the new orders component in particular. Fewer orders mean less activity in the coming months. The same goes for new durable goods orders, a separate data series focused on "big ticket" items like aircraft and washing machines. 4 | BUILDING PERMITS Don't forget about the housing market. "Housing IS The Business Cycle" is a famous paper by the economist Ed Leamer. That may not always be true, but they certainly tend to move up and down together. The first place trouble shows up is in building permits for new homes. You don't want to see these dropping at the same time all the aforementioned stuff is going on. If that's the case, it's a sign that a bigger, broader economic downturn may be coming. 5 | JOBLESS CLAIMS This is where the labor market should first start flashing warning signs. Yes, you can also track things like the Challenger job cut report and Indeed.com posting trends. But weekly new claims for jobless benefits have a very long and reliable track record. Just don't expect them to give you too much warning of a looming recession. During the last couple of downturns, claims only decisively moved up for about eight weeks or so before the recession hit. Still, they get bad before the monthly payroll data, so that does give you some lead time. 6 | "INDEX OF LEADING ECONOMIC INDICATORS" If this all feels like too much effort to follow, I have good news: The Conference Board has an "index of leading economic indicators" that includes most of what I just listed above. When you search for it, just make sure you look for the latest monthly release, which has more information than the website's generic landing page for the index. They're also not shy about including quotes that specifically tell you whether the data are pointing to a near-term recession. 7 | SIX MONTHLY INDICATORS HAVE PEAKED Now let's talk about how you know for sure that we're in recession. If you have to ask the question, it's probably too late. The rule of thumb is two consecutive quarters of contracting GDP. That's a decent gauge, but it means you might have to wait six months for the GDP data to confirm it. Plus, it's not technically accurate. The real definition is set by the National Bureau of Economic Research, and they look for signs that roughly six monthly indicators have peaked and are falling across the board. These include personal income, payrolls, household employment, real consumer spending, retail sales and industrial production. If these are all falling month after month, it's a bad sign. 8 | ST. LOUIS FEDERAL RESERVE ECONOMIC DATABASE If you're wondering "how on earth would I have time to check this all out," let me introduce you to FRED, the amazing St. Louis Federal Reserve economic database. Anytime I want to run a quick recession (or any other kind of data) check, I simply search for "st louis fred industrial production" and check the chart that's typically often the first or second result. If you want to get fancy, you can edit the charts to show year-on-year percent change and things like that. But, if you just take a quick glance at the overall levels, you'll get a feel for whether the economy is still expanding — or if it's plateaued or shrinking outright. 9 | MONTHLY PAYROLLS AND PERSONAL INCOME Back to the recession indicators. The NBER itself says its two most-important data points are the monthly payrolls and personal income numbers. So basically, it's not a full-blown recession until the labor market cracks. And by the way, once payrolls and income are shrinking for at least a month or two or three, unless there was a major natural disaster or something that pretty much guarantees, we're in a recession. Do not wait for headlines that say the NBER officially called it. They purposely wait until way after the fact so that once their declarations are made, they are super obvious to everyone, non-controversial, and unlikely to ever need to be changed. 10 | UNEMPLOYMENT RATE The reason I wanted to lay this all out sequentially is so that you know by the time we start talking about things like the unemployment rate, we almost already know for sure whether we're in recession. The unemployment rate is technically a lagging indicator, meaning it's one of the last places economic booms and busts show up. Anyone who's debating whether we're in recession and using things like the unemployment rate to make their case is kind of missing the point. The real debate should be had when the yield curve, jobless claims and new orders are all turning sour; it should be pretty much case closed by the time the unemployment rate starts clearly rising. 11 | INFLATION Another lagging indicator that's especially of relevance right now is inflation. Whether through the consumer price index (CPI), the Fed's preferred core personal consumption expenditures (core PCE), or any other inflation data point, there are definitely ways to slice and dice inflation to get a more leading trendline. Overall, however, prices in the economy tend to move very much with a lag. Think about it: Many fuel and food prices move more quickly, but things like hospital bills, airfare, restaurant meals all tend to reflect the labor market and the cost of workers, which is a lagging economic gauge in the first place. So, by the time conditions not only affect the labor market but then affect prices businesses charge for labor-intensive services, we could be a couple years after the fact. It doesn't make it feel any better to those coughing up higher prices, of course. But there is hope that if a recession is hitting, inflation should ultimately cave. It just takes some time.
