[0s - 120s]
Hello, hope you are doing well. So we are one week out-ish, a bit less, maybe six days out from the midterms in the U.S. So in this video, I wanted to kind of see if we can use AI to kind of predict the midterms, and I wanted to ask some more interesting questions that is a bit more related to kind of the markets and finance. How will this affect the S&P 500 the following day? So if the Republicans win or the Democrats win, we kind of wanna try to forecast what will happen, right? So for now, if you didn't know, I have my own startup, right? So this is kind of my AI superforecast startup, and I thought we can use that today to kind of see where we are looking towards the market because I've been running my model now for like, yeah, I think it's about four weeks, and this is the results on October 27th. We, our model has 90% on the House and 60% on the Senate, and we can compare this now. So our model is kind of the green line here, right? You can see we have been kind of following the market here on the House, but if we look at the Senate, we have kind of have this downwards trend lately. So we've been kind of, if we go into here, you can see we have been kind of doing this downwards trend. So we started here on round one. We were kind of following the market, but we were down. Then we kind of caught up to the market, but now we have this downwards trend again, and we are like six to eight points gap from Kalshi and Polymarket now. So I thought this was kind of an interesting, right? And we can also check how many seats we have predicted that the Democratic Senate will get. So we are at 51, while Kalshi is like 52. So we are pretty even on that, but and here we can see that yeah, every round we have, and now we are at the biggest gap. Now we are at six. So yeah, this is something I want to track, and it's interesting.

[120s - 240s]
For me to kind of keep to make this video, so I can go back and look at it when the when the elections is over next week. So like I said, we are gonna ask these questions, and I think we're just gonna get straight into that. So I went ahead and designed a couple of questions. So what we want to know now is if Democrats win control of the U.S. Senate on the November three midterm election, what will be the percentage change on the S&P 500 the next day, measured towards the November third closing price. So we're gonna do this in like eight bins, like I said, from minus two up to plus two, and or higher or lower, right? And we can basically do the opposite with the Republicans. So I did get like I had started right. You can see I started up my my Sift Intelligence custom questions with this, right? So basically, I thought we can just start another one in parallel, and if we switch out Democrats with Republicans, right? So we just do like this, switch it up here, and check question. And now our model is just gonna check everything, create the resolution rules, and we can get going on this. So while that spins up, let's take a take a look at some of the Senate races. So I think like kind of the most highlighted ones are kind of the Texas one, so Paxton versus Talarico. So our model now has this at like sixty thirty eight, and if you kind of look at the the development, let's find a graph here. You can see we have always our model has always been like a bit below the market, and we still are. We are like an eight point gap here between the market. We are still having Talarico as our favorite, but if we look at kind of the margins, our model estimate like a one point four win for Talarico. I think kind of the the Kalshi is basically almost the same, right? So let's look at another market.

[240s - 360s]
That we kind of this is kind of our biggest discrepancy from the market. So our model has Mary Peltola to get like fifty-fifty, almost like a toss-up. But the market is way above. So I've been kind of looking into this. Why does the AI kind of think Peltola is not more like a seventy-thirty? So you can see our models read is like history favors Sullivan. Alaska has not elected a Democratic senator since 2008. That's like a big one, right? And you can see the polls lean Peltola, what holds her back. Alaska polls have understated Republicans in recent elections. So basically, it's really hard to poll there, and it's more close to a coin flips, a coin flip, or like a toss-up is our models. So that's going to be interesting to track, I think, too. So if you just look kind of the full overview here, you can see we are pretty close to the market. We have a nine gap, nine-point gap in Maine. We have like a 17 in Alaska, as we said, a 14 in Nebraska, and other than that, we are pretty much on the market. So yeah, not so much big discrepancies from the market, but our model has picked some battles, and that's going to be interesting to see. And you can see we also track here how many points. This is for accuracy. So in Michigan, we have El-Sayed to win by four. In New Hampshire, we have Pappas to win by seven point six. Cooper to win by ten, right? And this is also we can something we can check when we actually have the results. This is basically just for my startup to kind of measure this, right? So let's take a look at now how we are doing on this. Yeah, we are still working on our two custom questions. So while we wait for that, let's take a look at how I've been doing because it's kind of fun to eat your own dog food. So on New Hampshire Senate state winner, we got in at 33. So this is up at 68 now. So this.

[360s - 480s]
Has been a double, so we have made hundred percent gains on this. That's been a really good bet on the house winner here. We put no on Brad Finstad. This has been going pretty well. And other than that, it's pretty much hasn't been the biggest moves. We are so I think the biggest downward swing is this long shot we did here. So we did Bobby Pulido to win the Texas fifteen house, but yeah, this was a long shot, right? Other than that, hasn't been much of movement. I guess this Ohio State Secretary winner hasn't done that well. We got in at sixty four. Now we are like a coin flip, so we have to wait and see on that. But I'm gonna place more positions when we get close to the to next week, I guess. So now I think we're just gonna wait for the two our two predictions on the S&P 500 and let's take a look at them. Okay, so now you can see we have the if the Democrats win the Senate or gonna win the control of the Senate, our model thinks that we will be on the slight negative side here at seventeen percent. If you look at all the bins, you can see we are spread quite thin here, so the model doesn't look that confident. But we can see some sign of that if the Democrats win, it will mostly be like from flat to yeah from flat, I guess, and downwards. It doesn't look too positive for the market if the Democrats win. And if you look at kind of how our model got there, you can see the questions ask the market moves the day after the midterm elections. A Democratic Senate under Trump most likely means a gridlock. So worries about AI stocks, possibly around strike, put it down. Sector worries in AI and risk of a post-election strike, right? And relief maybe when uncertainty clears. So this is kind of the reason why we the band holds to the middle and it doesn't look that positive for the market if the actually the Democrats take the Senate.

[480s - 600s]
But if you look at the other side, if the Republicans take the Senate, our model is a bit more positive on the S&P 500. So we landed on the 20% for a slight upwards trend. And if you kind of look at the spread here, here we are more on the upside. So we are like from flat up to like 1.5 is slightly above what we saw from the the Democratic win. So that is interesting, and that is something we can track next week. So if you kind of look at how model got here for the Republican, you can see a split Congress is seen as market friendly, right? A Republican Senate would be a surprise. We kind of know that now. A macro shocks could even put it lower, right? If the rate hike fears, oil nears high, and AI swing could also push push this downwards. Yeah, we we don't want to go through all of the reasoning here, but yeah, pretty interesting, and it's going to be interesting to see kind of next week where we land or where our model landed here. So if this is something you think is interesting, you can just go to the web page. I'm going to leave a link in the description, and you can start by just clicking the get early access. I'm probably going to open up a few more slots. It's been going really slow because I kind of need the space for the the API, so I got to kind of manage a few things. But I'm opening up slowly to more people to start using this. So you can just sign up on the waiting list here. This might look like there's not much you can do here, but there are actually a few use cases for this. Like I use it for some positions on maybe the market, but I'm going to go into I'm going to come into more use cases for this service in the future where I can actually take aboard a bit more people on it. So yeah, definitely go check it out if you're interested in stuff like that. So if you go here now, you can kind of see for today. I am actually waiting on the.

[600s - 684s]
Round 10 here now, so it's gonna be interesting to see when you can see the market has moved past our point here. But we will get our round 10 in today, and it's gonna be interesting to see if we follow this downward trend or if our model kind of bounces back up again towards the market. That should come in in like a few hours. So if you wanna follow that, you can. I'm gonna leave a link to the description to this page too, if you kind of wanna follow when I update this latest forecast here on the midterms election for the the house or was that the house? No, the Senate. Sorry. So yeah, that is basically my what I wanted to talk a bit about today, and I just wanted to document this because it could be interesting when I'm launching this startup to have something to look back at if we perform well. Like my expectations is pretty low, of course, because I kind of just started tracking. This is the first election I do. But yeah, if you like this kind of content, if you like like AI agentic trading and using like predictions using AI to try to place some position or just use it for like a starting place for a position, give this video a like and follow me because I'm gonna be doing videos like this. Yeah, the rest of the year, I guess. So yeah, thank you for tuning in today to my quick midterms update, and hopefully I see you again very soon.