More context would always be better. However, there is no reason Loess couldn't outperform a RandomForest and an SVM. Are we talking about just a single feature? If so, Random Forests would degrade into a bunch of jaggy trees. And I would assume a properly tuned SVM-regression would actually look quite similar to a Loess. If we're talking about multiple features, how exactly did you use Loess?

There is no perfect algorithm. I believe Loess, at least as implemented in R, is limited to ~4 features. Given so few features, the overhead of RandomForests or SVM-regression is likely wasted. It might be that the intrinsic scaling of the data is important and the RandomForest loses that in it's trees. For the SVM it could easily be the difficulty in properly tuning it or choosing the right kernel. If the relationship is simple enough, you don't need to expand in the faux-infinite dimensions of kernel space to understand it.

Having said that, just because Loess is better in this particular training set via cross-validation, that doesn't mean it will always be better. All models are just approximations.