The survey started 11 weeks ago
Eleven weeks ago we started to put out a weekly financial market survey. It has been mostly sent out via Twitter so far, and since I am active on FinTwit that is the bulk, but not all, of the people who have filled it out.
It took a few weeks to figure out the questions I wanted to include. I didn’t want to lead the respondent with the wording, and I didn’t want a survey that took 10 minutes either. In the end I settled with a survey consisting of nine total questions with two asking for typed out answers, six multiple choice questions, and then the final question asking what should change in the survey questions. It varies week to week but it seems to take about 90 seconds for the typical person to fill it out. One more important note is that for the weekend of 6/3/2022 I had a technical issue so there was no survey that week. For the numbers I just averaged the results from the previous and subsequent week, but the data is not “real.”
After the week of the 4th I plan on ramping up the survey by sending it out via Twitter, LinkedIn, Cold Email, Opt-In Email, and other socials. I have gotten as low as 9 and as high as 78 respondents by just sharing it on Twitter and retweeting it a few times. Clearly 100+ each week would be a big improvement and really the plan it to get around 1000/wk by the fall.
With all that said I think the results are interesting so let’s look at the questions, the results, and some sentiment tools we are starting to use.
The Fed has a NPS score
The First question is “Do you think the Federal Reserve is on the right track in regards to monetary policy? 10 being 100% and 0 being 0%.” The survey software I am using is called SurveySensum.com and it has a NPS (net promoter score) tool which I have been using for the 1-10 scale. So now we have a NPS score for the Fed lol. The follow up question asks “What is the primary reason for your score? And what should they be doing?” With that question we also get a wordcloud to show the words that pop up the most.
So with that lets look at a few graphics.
Here are the NPS dials for the question “Do you think the Federal Reserve is on the right track in regard to monetary policy?” As you can see the Fed did not have a single week where people approved of their performance. In fact the average NPS score over the last 10 weeks has been a dismal -32.53.
For the follow up question of “What is the primary reason for your score? And what should they be doing?” we got a wide variety of answers. For those we have a wordcloud week by week.
If you click through those wordclouds you might notice a few very consistent themes, most notably the word Inflation. I used another software platform, called Deep Talk, to analyze the aggregated results from the answers to this question. Among other tools it too has a wordcloud and then it has a sunburst chart. The Sunburst chart narrows down the data from the main topics and then subtopics. I will have more on this in future posts but for now the chart should be fairly straightforward.
As you can clearly see in the wordcloud below the top two topics, and it is not even close, were rates and inflation. This should not be a big surprise as that has been the dominant discussion for months now.
If we look at the Sunburst chart you can both see what the topic are and get a clue as to what people THINK the Fed SHOULD do to improve things. It would seem people think that the Fed needs to hike rates to fight inflation and they have been going too slow and need to wake up. Of course this is over the past 10 weeks, and in that time the Fed has shifted to this view with the most recent 75 bps hike. But referencing the week by week word cloud from earlier, it is clear that inflation is still the only thing people care about. In fact after a 75bps hike the Fed still got a horrible -37.14 NPS score.
Macro Market Forecasting
The next four questions in the survey are straightforward macro views. Each one asks “In 3 months where do you think the X will be?” and then you are offered 5-6 options. In the coming weeks (in 3 weeks to be exact) we will start to show the charts of how accurate the forecasts have been.
The first one is “In 3 months where do you think the US 10-Yr yield will be?” Unsurprisingly the responses have been very skewed to the upside. Looking at the average, and this is over the last 10 weeks, 23.08% of people thing rates will be up 50bps+ over the next 3-months, and 28.7% of people have thought that rates would be up 25-50bps over the next 3-months. Only 7.99% have thought long rates would drop 50bps+ over the next 3-months.
Next question is “In 3 months where do you think the SP500 will be?” Would you be surprised if I told you that people have gotten more bearish as price has dropped? During the weekend of 4/16/22 only 13.64% of people thought that the SP500 would be down -5% or more in the coming 3-months. Last weekend 6/17/22 42.86% though that the SP500 would be down -5% or more in the coming 3-months. And yes, last week the lowest rated answer was that the SP500 would be up 5%+ in the next 3-months with only 11.43% of people forecasting that. By the way, the SP500 was up over +6% in the past week.
Oil was up next. I went with WTI, but the same answers could be extrapolated to Brent for this purpose. The question is “In 3 months where do you think WTI Crude will be?” Instead of percentage or basis point terms I used dollars for crude, as that is how people always refer to it. As opposed to the previous questions the respondents seem to have forecast oil fairly well, I hope you were all selling vol. The most popular answers have been +/- $5, and +$5-15. So all of you thought that oil would be flat to slightly higher…and you have been right so far. Oil has gone up and then down, and is just above where we started 10 weeks ago. In three more weeks we can see how well the total estimates did, but it is looking good so far.
EUR/USD is the last market call. I almost went with the DXY, but that is a weak index and while I have a chart of it all day every day, I don’t really like it. EUR/USD on the other hand is a real currency pair. The question wording is “In 3 months where do you think the EUR/USD will be?” and we used pips as that is how FX is quoted. The results here have been interesting as the #1 answer has been 0. More accurately it has been -50-+50 pips. So basically unchanged. Over the 9-weeks this question has been on the survey it has averaged 31.98% of the answers, and really of the four market calls in the survey this one has had the most normal distribution with 200+ at 4.7%, +50-200 at 25.08%, -50-+50 31.98%, -50-200 28.49, and -200 and lower 9.75%. So far the EUR/USD has gone ever so slightly lower. The path has not been a straight line, but the results are not bad.
We will start to have several charts soon as we start hitting week-13 (3-months) and we can see what people forecasted and what actually happened. But so far with the little data we do have the classic go against the extremes has been more right than not. If there is a widespread extreme call then you might want to look at going the other way.
Household burn rate expectations
The next question in the survey is “Over the next 3 months do you expect to spend more or less than normal based on economic conditions? (will you be spending more or less) with 5=the same, 0=a lot less, and 10=a lot more.”
The answer to this question, essentially a household survey type question, has skewed every so slightly lower. if you add up the answers to 4 and 5 you get 57.02% of respondents expect to spend the same or slightly less over the next 3-months than they were doing beforehand. I get that the audience is limited and overrepresented by people on twitter in finance, but still despite this widespread talk of a recession very few people are planning to alter their household spend by very much at all.
Most underpriced risk
The last question we will look at here is “What do you think is the most underpriced risk in the market right now?” and the results have been…..well…consensus?
Look at the word cloud below. Inflation, rates, China, and recessions? I think most people would expect those to be top of mind the past year or so, and really always. But eventually I hope to unearth something that to me is an “unknown unknown,” but that day is not today.
As I have stated this is the early days of this data set. We are looking to scale it up quite a bit over the coming months, and in the future will do different surveys at all. Sentiment is one of the more interesting parts of the market, and from experience we know it can be very valuable. We plan on eventually releasing this data via an API so that others can access it on a weekly basis and use it in their own models. If you have any ideas or suggestions feel free to send them my way.
Happy Trading and Be Safe,
P.S. If you liked this then take a free two week trial of our service. If you have any questions send me an email or find me over at Twitter @DavidTaggart