## Related questions

# 🏅 Top traders

# | Name | Total profit |
---|---|---|

1 | Ṁ1,391 | |

2 | Ṁ1,286 | |

3 | Ṁ522 | |

4 | Ṁ361 | |

5 | Ṁ278 |

Seems quite likely according to CME https://www.cmegroup.com/markets/interest-rates/cme-fedwatch-tool.html

@weissz Assuming their predictions are well calibrated, this market should be at 97.4% (cf. Jun 12th meeting predictions)

@weissz At one point a March rate cut was 94%, so idk if CME is accurate. From looking at history, the Fed cuts later than everyone thinks, but once it cuts it cuts 100 bps at a time, not 25 bps.

@riverwalk3 I mean, this is just the literal market consensus (in that these probabilities are derived from the fed funds futures). Obviously predictions can be off, but one should expect this to be as well calibrated as it can be after conditioning on current info / lack thereof. Same exact principle as manifold, only involving actual money + analysts + firms for whom this is their whole job, as compared to us armchair analysts

@weissz Important to note that these probabilities are not actually tradable. The futures are very well priced, but they're just one number and could imply a range of probability distributions. So it's possible for the probabilities to be wrong while the futures are still right.

@dominic I wonder how they actually do the conversion to probabilities.

Here in Australia the overnight rate is always quite close to the Reserve Bank's target, which is a point and not a range. So when there are only two realistic possibilities (e.g. hold vs cut 25bps), you can extract the probability of each outcome from the futures prices (after doing a bit of calendar maths to get the periods the futures contracts apply to to line up with the Reserve Bank meetings) by assuming the price is a weighted sum of the two possible outcomes, there are no remaining degrees of freedom.

So extracting a probability for a cut at the next meeting is straightforward and likely to be well calibrated assuming the futures are well priced.

But the Fed's target is a range, and several meetings out the FedWatch tool lists probabilities for multiple outcomes. There's not enough information in the futures pricing alone to determine those numbers. Where do they get extra information from to solve this? Or is it simply based on historical data on the difference between futures prices and subsequent reality?

I guess I trust the probability for the next meeting, since I imagine that even though the target is a range, CME knows roughly where in the range the effective federal funds rate will likely be. Worst case they could just assume it would be at the same relative location within the bounds as it is now, that'd probably be good enough.

But I have wondered where the numbers come from.

@chrisjbillington All good Qs! They lay out the methodology here, but I haven’t yet read through in detail https://www.cmegroup.com/articles/2023/understanding-the-cme-group-fedwatch-tool-methodology.html

@chrisjbillington Even though the target is a range, the actual rate is nearly always 0.07-0.08 over the lower bound of the range (https://www.newyorkfed.org/markets/reference-rates/effr). Not sure why that is exactly but it seems to be true, so that's at least one less factor that would affect the prediction.