1: \begin{abstract}

2: The Az\'ema-Yor solution (respectively the Perkins solution) of the Skorokhod 

3: embedding problem has the 

4: property that it maximises (respectively minimises) the law of the 

5: maximum of the stopped process. We show that these constructions have a 

6: wider property in that they also maximise (and minimise) expected 

7: values for a more 

8: general class of bivariate functions $F(W_\tau,S_\tau)$ depending on the 

9: joint law of the 

10: stopped process and the maximum. Moreover, for monotonic functions 

11: $g$, they also maximise and 

12: minimise $\E[\int_0^\tau g(S_t) dt ]$ amongst embeddings of $\mu$, 

13: although, perhaps surprisingly, we show that for increasing $g$ the 

14: Az\'ema-Yor embedding minimises this quantity, and the Perkins embedding 

15: maximises it.

16: 

17: For $g(s)=s^{-2}$ we show how these results are useful in calculating model 

18: independent bounds on the prices of variance swaps.

19: 

20: Along the way we also consider whether $\mu_n$ converges weakly to $\mu$ 

21: is a sufficient condition for the associated Az\'ema-Yor and Perkins 

22: stopping times to

23: converge. In the case of the Az\'ema-Yor embedding, if the 

24: potentials at zero also converge, then the stopping times converge 

25: almost surely, {\cred but for the Perkins embedding this need not be the case. 

26: However, under a further condition on the convergence of atoms at zero, the 

27: Perkins stopping times converge in probability (and hence converge almost 

28: surely down 

29: a subsequence).} 

30: \end{abstract}

31: