Why Signs Matter at the Ends of the Real Line
A rational function can become arbitrarily large in magnitude as its input moves far to the right or left. Knowing only that its magnitude grows is not enough: the function may head toward positive infinity at one end and negative infinity at the other. Even when its limit is zero, its eventual sign tells us whether it approaches zero from above or below.
The leading terms of the numerator and denominator determine this end behavior. Their coefficients give the basic sign, while the degrees determine whether that sign changes between positive and negative inputs. The central task is to keep those two pieces of information separate and then combine them correctly.
Leading Terms and Eventual Sign
Let \(P\) and \(Q\) be nonzero polynomials, and suppose \(Q\) is not the zero polynomial. Write their leading terms as \(a x^m\) and \(b x^n\), where \(a\ne 0\), \(b\ne 0\), and \(m,n\) are nonnegative integers. The rational function is \(f(x)=P(x)/Q(x)\), wherever \(Q(x)\ne 0\).
For sufficiently large positive or negative \(x\), the denominator \(Q(x)\) is nonzero: its leading term dominates its lower-degree terms. Thus \(f(x)\) is defined far enough out in both directions, even if it is undefined at some finite inputs. The sign at an end is determined by the leading-term quotient \(C x^d\). If \(d\) is negative, this expression means \(C/x^{-d}\).
At positive infinity, \(x^d\) is positive for every integer \(d\), so the sign is the sign of \(C\). At negative infinity, the sign of the power depends on whether the degree difference is even or odd. If \(d\) is even, the sign is again the sign of \(C\); if \(d\) is odd, the sign is the opposite of the sign of \(C\). For a negative degree difference, parity has the same meaning as the parity of its absolute value.
The Degree Comparison Rule
The degree difference also tells us what happens to the magnitude of the function. A positive difference means the leading-term quotient grows without bound in magnitude. A zero difference means it approaches the coefficient ratio. A negative difference means it approaches zero.
- If \(d>0\), then \(f(x)\) tends to an infinite limit at each end. At \(+\infty\), the sign of that limit is the sign of \(C\). At \(-\infty\), it is the sign of \(C(-1)^d\).
- If \(d=0\), then \(\lim_{x\to+\infty}f(x)=\lim_{x\to-\infty}f(x)=C\).
- If \(d<0\), then \(\lim_{x\to+\infty}f(x)=\lim_{x\to-\infty}f(x)=0\). Its eventual sign at \(+\infty\) is the sign of \(C\), and its eventual sign at \(-\infty\) is the sign of \(C(-1)^{|d|}\). Thus these signs indicate the side from which the function approaches zero.
In the first case, a positive sign means the corresponding limit is \(+\infty\), and a negative sign means it is \(-\infty\). In the final case, the limit is zero regardless of sign: “approaches zero from below” does not mean the limit is negative.
Why the Rule Works
Factor out the leading power from each polynomial. Since all remaining terms have lower degree, their contributions relative to the leading term tend to zero as \(x\) tends to either end of the real line. We can therefore write
where \(u(x)\to 0\) and \(v(x)\to 0\) as \(x\to+\infty\) and as \(x\to-\infty\). Consequently, sufficiently far out, both \(1+u(x)\) and \(1+v(x)\) are positive. Their ratio is also positive and tends to \(1\). Dividing the factorizations gives
The final factor is eventually positive and tends to \(1\). It does not change the eventual sign, and it does not change whether the magnitude grows, tends to a nonzero constant, or tends to zero. The sign and magnitude behavior therefore come from \(Cx^d\). For positive \(x\), \(x^d>0\). For negative \(x\), its sign is determined by the parity of \(d\) (equivalently, of \(|d|\)). This proves the rule, including the signed conclusions at both ends.
Worked Examples
Worked Example: An Odd Degree Difference with Opposite Infinite Limits
Determine the signs and limits at both ends for
The leading coefficients are \(a=-6\) and \(b=2\), so \(C=a/b=-3\). The degrees are \(m=5\) and \(n=2\), so \(d=5-2=3\), which is odd and positive.
At \(+\infty\), \(x^3\) is positive, so \(Cx^3=-3x^3\) is negative and grows in magnitude. Thus \(\lim_{x\to+\infty}f(x)=-\infty\). At \(-\infty\), \(x^3\) is negative, so \(-3x^3\) is positive and grows in magnitude. Thus \(\lim_{x\to-\infty}f(x)=+\infty\).
To check the signs from the factored expression, divide by the leading terms:
Both the numerator and denominator of the fraction tend to \(1\), so the fraction is positive sufficiently far out in either direction. The expression therefore has the sign of \(-3x^3\) there, confirming both conclusions.
Worked Example: An Even Degree Difference with the Same Sign at Both Ends
Determine the signs and limits at both ends for
Here \(a=5\), \(b=-1\), \(m=6\), and \(n=2\). Hence \(C=5/(-1)=-5\) and \(d=6-2=4\). The difference is positive and even. At either end, \(x^4\) is positive and grows without bound, so \(-5x^4\) is negative and grows in magnitude. It follows that
Indeed, factoring gives
The fraction tends to \(1\), and its numerator and denominator are positive sufficiently far out on both sides. Thus the sign is eventually negative at both ends, as predicted. The denominator in the original function has leading term \(-x^2\), so it is also nonzero sufficiently far out.
Worked Example: Approaching Zero from Different Sides
Find the limits and eventual signs at both ends for
The leading coefficients are \(a=-4\) and \(b=2\), so \(C=-2\). The degree difference is \(d=2-5=-3\), which is negative and odd in parity. Therefore the limit is zero at both ends. At \(+\infty\), the eventual sign is the sign of \(C\), which is negative. At \(-\infty\), the odd power changes the sign, so the function is eventually positive.
The factorization makes both the size and the signs explicit:
The fraction tends to \(1\) and is positive sufficiently far out. Since \(x^{-3}=1/x^3\) tends to zero, \(h(x)\to 0\) at both ends. For positive \(x\), \(-2x^{-3}<0\); for negative \(x\), \(-2x^{-3}>0\). Thus \(h\) approaches zero from below at \(+\infty\) and from above at \(-\infty\), verifying the sign conclusions.
Worked Example: Equal Degrees Give the Same Finite Limit
Find the limits and eventual signs for
The degrees are both \(4\), so \(d=0\). The ratio of leading coefficients is \(C=-8/2=-4\). The theorem gives the same finite limit at both ends:
To check, divide numerator and denominator by \(x^4\):
As \(x\) tends to either end, the numerator tends to \(-8\) and the denominator tends to \(2\), so the quotient tends to \(-8/2=-4\). Since the limit is strictly negative, the function is eventually negative at both ends.
A Reliable Sign-Checking Process
When the expression is complicated, a short sequence of checks helps prevent errors. Record the leading coefficient ratio and degree difference before deciding either limit. Then use the parity of the difference only for the negative end; at the positive end, the power contributes no sign change.
Write \(a x^m\) for the numerator’s leading term and \(b x^n\) for the denominator’s.
Set \(C=a/b\) and \(d=m-n\). Keep the sign of \(C\) separate from the parity of \(d\).
If \(d>0\), the magnitude grows without bound; if \(d=0\), the limit is \(C\); if \(d<0\), the limit is zero.
At \(+\infty\), use the sign of \(C\). At \(-\infty\), keep that sign for an even difference and reverse it for an odd difference.
Common Mistake: Forgetting the Negative End
A frequent error is to use the sign of the leading coefficient ratio at both ends without considering the degree difference. For example, for \(p(x)=3x^5/(-x^2)\), the leading coefficient ratio is \(-3\). Concluding that \(p(x)\to-\infty\) at both ends uses that ratio but ignores the odd power difference.
In fact, \(p(x)=-3x^3\). When \(x\to+\infty\), this tends to \(-\infty\). When \(x\to-\infty\), \(x^3\) is negative, so \(-3x^3\) is positive and tends to \(+\infty\). The correction is to check the parity of \(m-n\) whenever \(x\) is negative. The coefficient ratio alone determines the sign only at positive infinity, or at both ends when the degree difference is even.
What to Keep in View
The degree comparison answers how large the rational function becomes; the leading coefficient ratio and parity answer which sign it has. These are related but distinct questions. In particular, a limit of zero has no positive or negative sign, even though the function may be eventually positive or negative as it approaches zero.
A useful final check is to compare your sign conclusion with the leading-term quotient \(Cx^{m-n}\). At the positive end, its sign is the sign of \(C\). At the negative end, its sign changes exactly when the degree difference is odd. If the degree difference is positive, these signs specify the corresponding infinite limits; if it is negative, they specify the side of zero.
Check Your Understanding
For each question, identify the leading coefficient ratio and degree difference before deciding the signs or limits.
- For \(f(x)=\frac{2x^7-1}{-x^4+3}\), what are the limits as \(x\to+\infty\) and \(x\to-\infty\)?
- For \(g(x)=\frac{-3x^6+2}{x^2+1}\), what is the eventual sign at each end?
- For \(h(x)=\frac{5x^2+1}{-2x^5+x}\), what is the limit at each end, and from which side does it approach zero?
- For \(r(x)=\frac{4x^3-x}{2x^3+1}\), what is the limit at each end?
- Why does an odd degree difference reverse the sign at negative infinity but not at positive infinity?
