I would be grateful for feedback on the class below. It is used to validate the cost and payment received:

public class Money     {         private readonly decimal _cost;         private readonly decimal _payment;         private readonly ICurrency _currency;          public Money(decimal cost, decimal payment, ICurrency currency)         {             if(cost<0)                 throw new ArgumentException("Cost cannot be less than zero", "Cost");             if (payment < 0)                 throw new ArgumentException("Payment cannot be less than zero", "payment");             if (payment < cost)                 throw new ArgumentException("Cost cannot be less than Payment Received");             if (decimal.Round(cost, 2) != cost)                     throw new ArgumentException(string.Concat("Cost has too many decimal places.  It should only have: ", currency.DecimalPlaces),"Cost");             if (decimal.Round(payment, 2) != payment)                 throw new ArgumentException(string.Concat("payment has too many decimal places.  It should only have: ", currency.DecimalPlaces), "payment");             if(currency ==null)                 throw new ArgumentNullException("Currency cannot be null", "ICurrency");             _cost = cost;             _payment = payment;             _currency = currency;         }           public decimal Cost         {             get { return _cost; }         }          public decimal payment         {             get { return _payment; }         }          public ICurrency Currency         {             get { return _currency; }         }          public static bool operator ==(Money money1, Money money2)         {             if (!ReferenceEquals(money1, null) &&                 ReferenceEquals(money2, null))             {                 return false;             }             if (ReferenceEquals(money1, null) &&                  !ReferenceEquals(money2, null))             {                 return false;             }             if (ReferenceEquals(money1, null) &&                  ReferenceEquals(money2, null))             {                 return true;             }             return money1.Equals(money2);         }          public static bool operator !=(Money money1, Money money2)         {             return !(money1 == money2);         }          public bool Equals(Money other)         {             if (ReferenceEquals(other, null))                 return false;             return _cost == other._cost && _payment == other._payment && _currency == other._currency;         }          public override bool Equals(object obj)         {             return Equals(obj as Money);         }          public override int GetHashCode()         {             unchecked             {                 int hash = 17;                 hash = hash * 23 + _cost.GetHashCode();                 hash = hash * 23 + _payment.GetHashCode();                 hash = hash * 23 + _currency.GetHashCode();                 return hash;             }         }     } 

The class is used by a Domain Service, which calculates the change required to meet a cost. Instead of doing this in the domain service:

public decimal CalculateChange(decimal cost, decimal payment, ICurrency currency) {    return payment-cost; } 

I can do this (with the code above):

public decimal CalculateChange(Money money) {    return money.payment-money.cost; } 

With the second approach; I can be sure that the cost and payment received values are validated. However, this seems like an enormous overkill for such a simple program. I would be grateful for any suggestions e.g. is there a good way the Domain Service can do the validation? Plus I would be grateful for feedback about the class in general.

I am trying to use the principle of least astonishment and Domain Driven Design. Therefore I would be grateful for answers in that context.