Recursive String Reverse

ReverseString.java

public class ReverseString {

  public static void main(String[] argv) {
    String origString = argv[0];
    System.out.println(origString);
    System.out.println(reverse(origString));
  }

  public static String reverse(String str) {
    if(str.length()<=1) return str;
    char[] strAry = str.toCharArray();
    return String.valueOf(reverseAry(strAry,0,strAry.length-1,strAry[0]));
  }

  public static char[] reverseAry(char[] strAry, int index, int end, char c) {
    strAry[index] = strAry[end];
    strAry[end] = c;
    return (++index<--end)?reverseAry(strAry, index, end, strAry[index]):strAry;
  }

}

If you return reverseAry(*) instead of assigning, you get tail recursion, which saves stack space. Also, for some reason, String.valueOf() is magically faster than new String() in many benchmarks. I have never done the research to figure out what the constructor does differently.