1: \begin{abstract}

2: 

3: We develop a duality theory for the problem of maximising expected

4: lifetime utility from inter-temporal wealth over an infinite horizon,

5: under the minimal no-arbitrage assumption of No Unbounded Profit with

6: Bounded Risk (NUPBR). We use only deflators, with no arguments

7: involving equivalent martingale measures, so do not require the

8: stronger condition of No Free Lunch with Vanishing Risk (NFLVR). Our

9: formalism also works without alteration for the finite horizon version

10: of the problem. As well as extending work of Bouchard and Pham

11: \cite{bp04} to any horizon and to a weaker no-arbitrage setting, we

12: obtain a stronger duality statement, because we do not assume by

13: definition that the dual domain is the polar set of the primal

14: space. Instead, we adopt a method akin to that used for inter-temporal

15: consumption problems, developing a supermartingale property of the

16: deflated wealth and its path that yields an infinite horizon budget

17: constraint and serves to define the correct dual variables. The

18: structure of our dual space allows us to show that it is convex,

19: without forcing this property by assumption. We proceed to enlarge the

20: primal and dual domains to confer solidity to them, and use

21: supermartingale convergence results which exploit Fatou convergence,

22: to establish that the enlarged dual domain is the bipolar of the

23: original dual space. The resulting duality theorem shows that all the

24: classical tenets of convex duality hold. Moreover, at the optimum, the

25: deflated wealth process is a potential converging to zero. We work out

26: examples, including a case with a stock whose market price of risk is

27: a three-dimensional Bessel process, so satisfying NUPBR but not NFLVR.

28: 

29: \end{abstract}

30: