WEBVTT - Here's Why Company Earnings Are So Difficult to Forecast

0:00:02.440 --> 0:00:06.760
<v Speaker 1>Bloomberg Audio Studios, Podcasts, radio news.

0:00:08.360 --> 0:00:10.959
<v Speaker 2>I'm Stephen Carroll and this is Here's Why, where we

0:00:11.000 --> 0:00:13.119
<v Speaker 2>take one news story and explain it in just a

0:00:13.119 --> 0:00:20.520
<v Speaker 2>few minutes with our experts Here at Bloomberg, earning season

0:00:20.680 --> 0:00:24.400
<v Speaker 2>can feel like marking a scorecard. We compare the numbers

0:00:24.440 --> 0:00:28.560
<v Speaker 2>companies reports to what they and industry experts were expecting.

0:00:29.000 --> 0:00:31.040
<v Speaker 1>Another big day on the earnings from then Reno and

0:00:31.080 --> 0:00:34.080
<v Speaker 1>this is crucial maintaining its full year guidance.

0:00:34.159 --> 0:00:37.000
<v Speaker 2>That's after some of its rivals think of Stalantis, think

0:00:37.000 --> 0:00:39.440
<v Speaker 2>of BMWVW Cutts. Therefore you guidance.

0:00:39.520 --> 0:00:42.120
<v Speaker 1>General Motors are up about two percent posted better than

0:00:42.159 --> 0:00:43.400
<v Speaker 1>expected results.

0:00:43.120 --> 0:00:44.400
<v Speaker 2>Or mess numbers coming out here.

0:00:44.520 --> 0:00:47.320
<v Speaker 1>It looks like their sales are actually beating the estimates.

0:00:47.320 --> 0:00:48.720
<v Speaker 1>Remember this is coming right off of the.

0:00:48.680 --> 0:00:50.840
<v Speaker 2>Heels of caring in the last twenty four hours, which

0:00:50.880 --> 0:00:54.160
<v Speaker 2>showed the exact opposite. These reports give us a look

0:00:54.240 --> 0:00:57.480
<v Speaker 2>under the hood whether business is good or not, and

0:00:57.600 --> 0:00:59.840
<v Speaker 2>also give us an insight into what companies are seeing

0:01:00.000 --> 0:01:04.400
<v Speaker 2>coming down the line. Meeting, beating or missing expectations are

0:01:04.440 --> 0:01:07.759
<v Speaker 2>some of the most important metrics for investors, but it's

0:01:07.800 --> 0:01:11.280
<v Speaker 2>not easy to pick exactly where to set your expectations.

0:01:11.959 --> 0:01:18.880
<v Speaker 2>Here's why company earnings are so difficult to forecast. To

0:01:18.959 --> 0:01:21.480
<v Speaker 2>explain how it's done, I've got Gena Martin Adams with me,

0:01:21.560 --> 0:01:25.240
<v Speaker 2>chief equity strategist for Bloomberg Intelligence, Gina, great to have you.

0:01:25.240 --> 0:01:28.199
<v Speaker 2>Can you give us a peek inside your crystal ball?

0:01:28.480 --> 0:01:30.520
<v Speaker 2>What sort of things do you look at when deciding

0:01:30.720 --> 0:01:32.640
<v Speaker 2>what to expect from a company's report?

0:01:32.920 --> 0:01:35.760
<v Speaker 1>Yeah? Great question. The crystal ball is often very cloudy

0:01:35.760 --> 0:01:38.160
<v Speaker 1>when it comes to earnings, in particular, when you're taking

0:01:38.360 --> 0:01:43.080
<v Speaker 1>five hundred companies estimates, aggregating them to an index level,

0:01:43.440 --> 0:01:46.160
<v Speaker 1>and then trying to determine if the degree of accuracy

0:01:46.400 --> 0:01:49.920
<v Speaker 1>becomes a little bit suspicious, you know, I think thats

0:01:50.240 --> 0:01:53.480
<v Speaker 1>what most forecasters do is look at the consensus and

0:01:53.880 --> 0:01:58.600
<v Speaker 1>run through their models how their expectations might deviate from consensus.

0:01:58.600 --> 0:02:03.160
<v Speaker 1>The consensus, of course, is the community of other analysts,

0:02:03.720 --> 0:02:06.640
<v Speaker 1>which I think ultimately creates a lot of error in

0:02:06.880 --> 0:02:11.080
<v Speaker 1>company analysis because ultimately the consensus should be the market,

0:02:11.720 --> 0:02:15.320
<v Speaker 1>and the market is not always priced exactly in line

0:02:15.360 --> 0:02:18.520
<v Speaker 1>with how analysts are expecting earnings to play out. And

0:02:18.520 --> 0:02:21.440
<v Speaker 1>we see this quite often through the options markets, which

0:02:21.480 --> 0:02:25.040
<v Speaker 1>probably provide us with a greater degree of accuracy than

0:02:25.160 --> 0:02:30.119
<v Speaker 1>analysts forecasts do over time because of participation. There are

0:02:30.200 --> 0:02:35.079
<v Speaker 1>just so many people betting on earnings anticipating earnings growth

0:02:35.160 --> 0:02:38.160
<v Speaker 1>that the consensus view of the market tends to be

0:02:38.240 --> 0:02:40.400
<v Speaker 1>more correct. But we do a lot of work on this.

0:02:40.520 --> 0:02:45.680
<v Speaker 1>We assess analyst consensus, compare that to company guidance, Compare

0:02:45.720 --> 0:02:49.119
<v Speaker 1>that to what's priced in the market. It's different across

0:02:49.240 --> 0:02:52.560
<v Speaker 1>all industries, it's different across all companies, it's different across

0:02:52.560 --> 0:02:55.400
<v Speaker 1>all sectors, and then we aggregate it to the index

0:02:55.520 --> 0:02:58.440
<v Speaker 1>level to try to get an assessment of what's anticipated

0:02:58.840 --> 0:03:04.240
<v Speaker 1>and what might ultimately beat or meet or miss those expectations.

0:03:04.639 --> 0:03:07.200
<v Speaker 2>It's a complex task and a massive one that you

0:03:07.280 --> 0:03:11.200
<v Speaker 2>take on every earning season. I wonder does the disclaimer

0:03:11.280 --> 0:03:15.160
<v Speaker 2>you know, past performance does not guarantee future returns relevant

0:03:15.240 --> 0:03:19.000
<v Speaker 2>when it comes to these source of expectations. Is it

0:03:19.080 --> 0:03:21.600
<v Speaker 2>that you look at what's gone before and try to

0:03:21.600 --> 0:03:24.400
<v Speaker 2>extrapolate what's happening next or talk us through a bit

0:03:24.440 --> 0:03:25.440
<v Speaker 2>of the science involved.

0:03:25.600 --> 0:03:29.600
<v Speaker 1>Yeah. So oftentimes what happens with the analyst community anyway,

0:03:29.639 --> 0:03:32.480
<v Speaker 1>at the individual stock level is they run a company model,

0:03:33.280 --> 0:03:37.000
<v Speaker 1>and that company model is based upon past relationships as

0:03:37.040 --> 0:03:41.200
<v Speaker 1>well as anticipated future market or market share changes, future

0:03:41.280 --> 0:03:45.160
<v Speaker 1>cost input changes. There's a lot that goes into each

0:03:45.200 --> 0:03:48.160
<v Speaker 1>individual analyst forecast, and then what we do is we

0:03:48.240 --> 0:03:51.560
<v Speaker 1>aggregate those analysts forecasts to an index level and get

0:03:51.560 --> 0:03:55.000
<v Speaker 1>an assessment of what is anticipated by the analyst community

0:03:55.120 --> 0:03:58.560
<v Speaker 1>for all stocks in a given index. The other way

0:03:58.600 --> 0:04:01.240
<v Speaker 1>that you can do this we definitely do this in

0:04:01.280 --> 0:04:03.680
<v Speaker 1>our team as well, is look at company guidance and

0:04:04.560 --> 0:04:09.800
<v Speaker 1>use historical guidance as an indication of what is likely

0:04:09.840 --> 0:04:12.560
<v Speaker 1>to come in the earning season. For example, the last

0:04:12.560 --> 0:04:16.200
<v Speaker 1>two earning seasons are guidance based model, which really just

0:04:16.240 --> 0:04:20.719
<v Speaker 1>takes what companies say is likely has anticipated much stronger

0:04:20.760 --> 0:04:23.679
<v Speaker 1>growth than the analyst community has anticipated and has actually

0:04:23.720 --> 0:04:26.800
<v Speaker 1>given us some pretty good direction as to what to expect.

0:04:27.279 --> 0:04:29.160
<v Speaker 1>And then the third way that we look at the

0:04:29.360 --> 0:04:32.719
<v Speaker 1>broader market is also to use macro inputs to try

0:04:32.720 --> 0:04:35.200
<v Speaker 1>to forecast earnings. We don't do this necessarily on a

0:04:35.240 --> 0:04:38.200
<v Speaker 1>quarterby quarterbasis, but really do this looking out over the

0:04:38.200 --> 0:04:40.080
<v Speaker 1>next twelve months, what are the moving parts of the

0:04:40.120 --> 0:04:43.880
<v Speaker 1>macro anticipate for earnings? And these are very very simple

0:04:43.880 --> 0:04:48.240
<v Speaker 1>regressions where we just look at typical what has historically

0:04:49.080 --> 0:04:52.400
<v Speaker 1>proven indicative of earnings from a macro perspective. Use those

0:04:52.440 --> 0:04:55.560
<v Speaker 1>inputs to try to estimate what's happened what's likely to

0:04:55.560 --> 0:04:58.760
<v Speaker 1>happen with earnings over the next twelve months. This is

0:04:59.440 --> 0:05:03.280
<v Speaker 1>actually proven to be a pretty effective way of forecasting

0:05:03.320 --> 0:05:06.400
<v Speaker 1>long term earning strends. It's been terrible the last two

0:05:06.520 --> 0:05:08.760
<v Speaker 1>years as the index, especially in the S and P

0:05:08.839 --> 0:05:12.880
<v Speaker 1>five hundred, has deviated materially from the macro economy. Given

0:05:13.520 --> 0:05:17.960
<v Speaker 1>this abnormally strong performance and earnings leadership of the mag

0:05:18.080 --> 0:05:23.040
<v Speaker 1>seven which has created this very strong earnings recovery absent

0:05:23.279 --> 0:05:25.920
<v Speaker 1>a very strong macro backdrop.

0:05:25.640 --> 0:05:28.800
<v Speaker 2>I wonder who gets it most right of everyone that's

0:05:28.839 --> 0:05:31.520
<v Speaker 2>looking and trying making these forecasts. Do the analyst community

0:05:31.520 --> 0:05:34.600
<v Speaker 2>tends be closer to what happens? Or are companies themselves

0:05:34.720 --> 0:05:37.280
<v Speaker 2>generally a good guide to what actually transpires.

0:05:37.520 --> 0:05:40.680
<v Speaker 1>Yeah, it's a great question, and the short answer is

0:05:40.680 --> 0:05:45.840
<v Speaker 1>that changes. So sometimes. The macro is a great forecast tool.

0:05:45.920 --> 0:05:48.840
<v Speaker 1>As I mentioned, historically, we've got a very high R

0:05:48.880 --> 0:05:51.440
<v Speaker 1>squared in our regression model. It exists for a reason.

0:05:51.480 --> 0:05:54.320
<v Speaker 1>It does a very good job of anticipating where earnings

0:05:54.320 --> 0:05:57.520
<v Speaker 1>are likely to head normally, but we've been in an

0:05:57.560 --> 0:06:03.640
<v Speaker 1>abnormal environment. Analysts have gotten less precise over time. I

0:06:03.839 --> 0:06:07.440
<v Speaker 1>can't explain exactly why that has happened, but they've become

0:06:07.560 --> 0:06:11.560
<v Speaker 1>in particular, over the last three years almost perma bears

0:06:11.839 --> 0:06:16.960
<v Speaker 1>permanently underestimating company potential for earnings growth. It's difficult to

0:06:17.000 --> 0:06:20.040
<v Speaker 1>determine if that's because of this macro divergence that has

0:06:20.080 --> 0:06:23.839
<v Speaker 1>come out or not. And recently guidance has given much

0:06:23.880 --> 0:06:26.800
<v Speaker 1>better indication of what's likely to come for the index

0:06:26.839 --> 0:06:30.160
<v Speaker 1>at large. Now, the most interesting thing about guidance is

0:06:30.400 --> 0:06:36.080
<v Speaker 1>it's pretty anomaloust to see such fantastic accuracy emerge from

0:06:36.080 --> 0:06:39.039
<v Speaker 1>the guidance numbers. And I say that because only about

0:06:39.040 --> 0:06:41.880
<v Speaker 1>a fifth of S and P five hundred companies give

0:06:41.960 --> 0:06:46.560
<v Speaker 1>us guidance at any point in time, and therefore we're

0:06:46.800 --> 0:06:51.760
<v Speaker 1>relying on just twenty percent of companies guiding on expectations

0:06:51.800 --> 0:06:56.040
<v Speaker 1>to drive our expectation for the broad market. But nonetheless

0:06:56.200 --> 0:06:59.719
<v Speaker 1>their guidance has proven to be very accurate and quite

0:06:59.760 --> 0:07:03.640
<v Speaker 1>aive of earnings trends, at least over the last few years.

0:07:03.920 --> 0:07:06.400
<v Speaker 2>That's so interesting because I did wonder. I mean, obviously,

0:07:06.480 --> 0:07:09.600
<v Speaker 2>everyone wants to surpass expectations when it comes to the

0:07:09.920 --> 0:07:12.160
<v Speaker 2>reported numbers. But I did wonder if there was ever

0:07:12.200 --> 0:07:14.880
<v Speaker 2>a trend of companies under promising so that they can

0:07:14.920 --> 0:07:16.160
<v Speaker 2>then over deliver.

0:07:16.520 --> 0:07:19.800
<v Speaker 1>Yeah, I think that that has historically been the case,

0:07:20.160 --> 0:07:24.560
<v Speaker 1>and there's still somewhat under promising, but the analyst community

0:07:24.600 --> 0:07:30.160
<v Speaker 1>has still been so pessimistic and undershooting reality that guidance,

0:07:30.200 --> 0:07:33.080
<v Speaker 1>even though they're underpromising and over delivering, their under promises

0:07:33.640 --> 0:07:35.640
<v Speaker 1>are still above analyst consensus.

0:07:36.000 --> 0:07:36.360
<v Speaker 2>Wow.

0:07:36.640 --> 0:07:39.080
<v Speaker 1>And I think that that's largely down to tech and

0:07:39.120 --> 0:07:42.480
<v Speaker 1>this phenomenon that has occurred in the tech industry with

0:07:42.520 --> 0:07:45.920
<v Speaker 1>respect to AI and the mag seven earners. When we

0:07:45.960 --> 0:07:49.080
<v Speaker 1>look at guidance, the vast majority of guidance comes from

0:07:49.440 --> 0:07:53.240
<v Speaker 1>tech and consumer discretionary sectors anyway. So we get a

0:07:53.240 --> 0:07:56.200
<v Speaker 1>really good feel from companies on what's happening with the

0:07:56.240 --> 0:07:58.400
<v Speaker 1>consumer outlook, the global consumer out look. We get a

0:07:58.400 --> 0:08:00.840
<v Speaker 1>pretty good feel from companies what's happening with global tech,

0:08:01.600 --> 0:08:06.320
<v Speaker 1>and those industries have been quite dominant in the index.

0:08:06.520 --> 0:08:09.240
<v Speaker 1>So I think that's what's happening. But it is fascinating

0:08:09.280 --> 0:08:13.800
<v Speaker 1>to see how analysts really just have not jumped on

0:08:13.880 --> 0:08:17.040
<v Speaker 1>the bandwagon of optimism, even though companies continually tell them

0:08:17.080 --> 0:08:18.240
<v Speaker 1>they should be more optimistic.

0:08:18.480 --> 0:08:21.120
<v Speaker 2>When we think about how the reaction comes to the

0:08:21.640 --> 0:08:24.640
<v Speaker 2>reports as we get them, is there room for nuance?

0:08:24.800 --> 0:08:27.720
<v Speaker 2>Can you beat on one metric, disappoint on another and

0:08:27.960 --> 0:08:30.120
<v Speaker 2>kind of end up with a positive share price reaction

0:08:30.160 --> 0:08:30.560
<v Speaker 2>in the end?

0:08:30.720 --> 0:08:33.520
<v Speaker 1>Oh? Absolutely, And you can also have the reverse and

0:08:33.559 --> 0:08:37.079
<v Speaker 1>that ends up being where the options market really comes

0:08:37.120 --> 0:08:42.080
<v Speaker 1>in handy. Is oftentimes what's priced in a stock is

0:08:42.120 --> 0:08:47.160
<v Speaker 1>not actually reflected in analyst expectations or guidance at all.

0:08:47.200 --> 0:08:49.160
<v Speaker 1>And we saw this emerge a little bit in the

0:08:49.200 --> 0:08:52.679
<v Speaker 1>second quarter. If you recall, going into that July reporting season,

0:08:52.800 --> 0:08:55.960
<v Speaker 1>expectations were very high for tech companies at large, in

0:08:56.000 --> 0:09:00.560
<v Speaker 1>particular AI related industries. Across the board. Tech commpanies beat

0:09:00.600 --> 0:09:04.880
<v Speaker 1>those expectations. They even guided for higher growth going forward,

0:09:04.920 --> 0:09:07.920
<v Speaker 1>but in many cases the stocks did not perform particularly

0:09:07.960 --> 0:09:11.600
<v Speaker 1>well in response to earnings. And that's because there was

0:09:11.640 --> 0:09:14.840
<v Speaker 1>a bunch of nuance in the detail. And that detail

0:09:15.600 --> 0:09:18.880
<v Speaker 1>really with that sector was a regarding margin and the

0:09:18.920 --> 0:09:23.280
<v Speaker 1>potential sustainability of margin recovery. That has emerged as a

0:09:23.280 --> 0:09:25.679
<v Speaker 1>powerful driver of earnings growth in that segment for the

0:09:25.760 --> 0:09:28.560
<v Speaker 1>last year and a half or so, a company started

0:09:28.559 --> 0:09:31.880
<v Speaker 1>talking about, well, we're still going to beat expectations, we

0:09:31.920 --> 0:09:34.240
<v Speaker 1>still see very strong growth, but we might have to

0:09:34.280 --> 0:09:36.280
<v Speaker 1>spend a little bit more than we had anticipated. We're

0:09:36.280 --> 0:09:40.440
<v Speaker 1>really adjusting our spending plans, we're really assessing digging a

0:09:40.440 --> 0:09:42.920
<v Speaker 1>bit deeper into the spending outlook, and that created a

0:09:42.920 --> 0:09:46.040
<v Speaker 1>lot of nervousness in the market. So you can absolutely

0:09:46.080 --> 0:09:49.120
<v Speaker 1>have these key issues that emerge that are oftentimes very

0:09:49.200 --> 0:09:54.400
<v Speaker 1>nuanced or oftentimes extremely detailed. Even though they beat expectations,

0:09:54.600 --> 0:09:58.319
<v Speaker 1>raised guidance, generally had a very positive earning season, that

0:09:58.440 --> 0:10:02.040
<v Speaker 1>commentary can so drive stock prices. This is a big

0:10:02.200 --> 0:10:05.080
<v Speaker 1>part of the reason why we also track sentiment. We

0:10:05.120 --> 0:10:08.960
<v Speaker 1>have a transcript analysis tool that we've developed. We track

0:10:09.080 --> 0:10:13.880
<v Speaker 1>sentiment toward keywords, sentiment in management commentary as well as

0:10:13.880 --> 0:10:18.280
<v Speaker 1>in analyst questions toward keywords because you can often get

0:10:18.440 --> 0:10:21.760
<v Speaker 1>enormous stock price reactions and a lot of signals from

0:10:21.800 --> 0:10:23.760
<v Speaker 1>the words and the reactions to the words.

0:10:24.040 --> 0:10:27.439
<v Speaker 2>Gina Mark Adam's chief equity strategist a Bloomberg Intelligence, thank

0:10:27.440 --> 0:10:29.440
<v Speaker 2>you so much for joining us and giving us an

0:10:29.440 --> 0:10:32.920
<v Speaker 2>insight into how these sorts of forecasts are put together

0:10:33.000 --> 0:10:35.319
<v Speaker 2>for more explanations like this from our team of twenty

0:10:35.320 --> 0:10:37.800
<v Speaker 2>seven hundred journalists and analysts around the world. Search for

0:10:37.880 --> 0:10:40.720
<v Speaker 2>quick take on the Bloomberg website or Bloomberg Business app.

0:10:42.160 --> 0:10:44.680
<v Speaker 2>I'm Stephen Carol. This is here's why. I'll be back

0:10:44.720 --> 0:10:46.480
<v Speaker 2>next week with more. Thanks for listening.